Hypertrophy training – It’s not that complicated

Hypertrophy training – It’s not that complicated

I’ve included a TLDR summary for this article. Because each individual subsection is a little inconclusive (because there is a massive grey zone in most respects where a huge number of approaches are viable) you probably won’t get the point until you’ve read the lot.

So I suggest reading the TLDR before you begin even if you intend to soldier through the entire article because it will help you put the bits in context. Maybe.

The TLDR version

1 – It is fairly conclusive from the literature that higher weekly volumes lead to greater hypertrophy gains. It also seems that the distribution of that volume is much less important, with both very low and very high training frequencies being viable. That said, I think there is both a theoretical and practical basis to assume that training with moderate per-session volumes around twice a week is best for hypertrophy most of the time.

2 – The ultimate driver of hypertrophy chronically is not volume, or frequency, but the application of sufficient training stress chronically to promote adaptation, of which volume/frequency arrangements describe some part. Attempts to maximise the hypertrophic stimulus (through increased volume) or maximise frequency need to be placed in this context. Evidence of well-applied training stress will be apparent in your ability to progressively overload. 

3 – You should adopt training structures that allow you to apply sufficient stress to elicit adaptation and recover sufficiently to do so again on a chronic basis. For this reason hypertrophy training benefits from more strategic planning than many people give it.

4 – The ULTIMATE TLDR – while there is probably an optimal volume and frequency dose, these variables are just methods of controlling training stress and the ultimate proof of their efficacy is in the outcomes (improved performance in relevant measures, bigger muscles) and not how closely they conform to a theoretical “optimal”.

Addendum 2019
Since I’ve written this article I have changed/refined my opinions on a few things. I’ll eventually get around to fleshing out those thoughts further, but for now, in addition to the below, you can check out
Another article that I wrote discussing per-session volumes, failure proximity, and exercise selection
A podcast with my colleague Luke Tulloch where we discuss the relationships between volume, strength and hypertrophy

At the bottom of the article I’ve also included some further reading from other authors that covers some of the same ground that I do, and either supports or fleshes out some of the points I made further.


The title is deliberately ironic. As it happens, hypertrophy training is in theory very complicated and in practice pretty simple. This is actually more of an opinion piece than an attempt to really synthesise the literature, so take it as such – I have a philosophy for how people ought to train for the majority of time for size and basically went about collating and reading the evidence I needed to put it forward, then cited it very lazily and sparingly. That my opinion has some evidence backing it isn’t entirely incidental, of course, but I’m more than open to the fact that I might be wrong on a few fronts so let me know.
Also, there’s a bunch of concepts in here (repeated bout effect, size principle, motor units, micro/meso/macrocycles etc) that I don’t define/explain, and I’d suggest googling them if you aren’t familiar with them, as it’ll help make sense of the rest of this.

One of the great upsides of the increasing focus on science in strength sports has been the improved conceptual understanding of the role of individual training variables in promoting adaptations, coupled with increased mechanistic knowledge of the actual processes of getting jacked(er) and strong(er).

However, transposing research to practice remains difficult. Training interventions are multifaceted and variables that are deliberately explored in isolation in research are interdependent. A myopic focus on one optimising one variable may come at the expense of a more appropriate program when we look at things more globally.

Which brings me neatly to volume, frequency, and hypertrophy training. Although I think the role of both volume and frequency are better understood than previously, and although I think we have relatively good recommendations for building hypertrophy programs, I also think those are lost in the noise a little bit, with people relying on soundbite science (eg “volume drives hypertrophy”, or “protein synthesis returns to baseline after only a couple of days”) to draw conclusions about effective training that lead them down the garden path.

So what I’d like to do in this article is put a little meat on the bones of those thoughts and suggest some effective hypertrophy training practices with consideration to both.

What is hypertrophy and why does it happen?

 Hypertrophy is an increase in the size of muscle tissue. This encompasses increases in the contractile elements of a muscle (ie the fibers) and a concomitant expansion of the extracellular matrix. When we lift weights a whole bunch of jiggery-pokery (that’s a technical term used in molecular biology) results in the addition of contractile units in parallel, leading to an increased diameter per fiber and thus an increase in muscle cross sectional area.
Hypertrophy CAN also happen in series to adapt to a new functional length (eg in the case of previously immobilized limbs) and as a response to eccentric training. While that might be important for injury prevention and rehab, it’s not really the focus of people getting swole so we won’t get too into that. Similarly, whilst hyperplasia (the addition of new fibers, as opposed to growing existing ones) has been demonstrated in some animal models, it doesn’t appear to happen with training. Off the top of my head there was some research showing a greater total number of muscle fibers in bodybuilders than regular folk, but that didn’t demonstrate that the disparity arose as a result of training per-se, rather that bodybuilding might just select for those with higher than average muscular potential.

Of some importance, also, is the idea of a “myonuclear domain”. Muscle tissues are multinucleated – the little powerhouses that actually synthesise new proteins etc are distributed along the long muscle cells. Myonuclei can only oversee a given amount of cell volume, and so eventually for growth to occur additional nuclei have to be added.
So in addition to the aforementioned jiggery-pokery that allows us to synthesise new muscle tissue, it’s presumed that at some point you need satellite cells to differentiate into additional nuclei to facilitate this, and researchers have found more or less this, with only modest growth possible without the addition of said myonuclei (9).
Quite cool (and again quite irrelevant to the article at hand) is that additional myonuclei appear to stay around even when we atrophy (lose size), which facilitates regain of muscle size (so-called muscle memory). I recently learnt that there’s a bit more to it than that, with some cool epigenetic stuff going on too, but the point is you need enough little muscle-making factories distributed along your muscle fibers to grow and I’ll leave it there.

So hypertrophy is the addition of contractice and non-contractile elements to muscle tissue. In response to training it happens mostly by the addition of contractile elements in parallel. The question is WHY/what causes it (“Why” seems a little too metaphysical of a question for a meathead like me to answer).

As with everything, our adaptations are fundamentally a response to imposed stressors. We have a baseline level of muscularity, are existing in a nice little homeostatic comfort zone, suffer a perturbation (training), and respond to that by fortifying the systems that were stressed. There is a threshold of stress required to actually induce adaptation, and as the systems get increasingly resilient that threshold increases.

In muscle hypertrophy, 3 major drivers of growth have been identified.

1 – Muscle Damage – this one is actually the most disputed. I’ll outline the reasons it may be beneficial below, but people have observed that muscle growth can occur in training involving minimal muscle damage, and that early size increases observed in programs can reflect oedema (swelling) as opposed to increased muscle protein content, with hypertrophy occurring after this (1). Because muscle protein synthesis (MPS) can occur both to repair damage and to lay down new tissue, assuming that highly damaging bouts of exercise increase growth due to greater increases in MPS, reflecting GROWTH per se, may be misleading. Additionally, muscle damage causes reduced contractile function until it is repaired, limiting the frequency with which training can occur.

On top of that, it is very hard to examine muscle damage in the absence of the application of tension (which we’ll also discuss), so distinguishing its unique effects is hard, at the least. Somebody induced contusion injuries on rats in 2001, which I’ve heard people bring up in this discussion, observing a loss of some fibers and an increase in the size of those that remained, but whether you can extrapolate that to exercise-induced damage in humans, and what that would mean anyway is a little beyond me (2).

One thing that does seem logical is that if early increases in MPS reflect repair of damaged tissue, it would follow that the repeated bout effect (reduction in disruption from training that you have acclimated to) might actually help, as it would allow greater net application of training stress, which makes a case for limiting the frequency with which you introduce novel exercises and loading schemes, and instead focus on improving performance in the ones you are using for a reasonable time period, but we’ll get to that much later too.

So if muscle damage may NOT be helpful, it also MAY be helpful for a couple of reasons. Firstly, in the repair of cellular debris, signalling molecules are produced that have downstream effects promoting hypertrophy. Secondly, damage promotes the activation and differentiation of satellite cells (remember I mentioned them earlier), which can donate additional myonuclei (remember them now?) and also produce some growth/repair promoting factors (3)

Which is to say, some muscle damage might have some benefits, but it’s hard to say if it’s truly NECESSARY, and given that it is hard to distinguish muscle damage from stress induced by tension (noting that things such as higher volumes and long ranges of motion increase damage – 4), I don’t think chasing muscle damage in the bulk of your training is a great idea. That also said, given that optimizing tension and metabolic stimuli (the next bits) probably entails inducing some damage, I think avoiding it at all costs mightn’t be the best idea either, so maybe just consider it incidental.

2 – Mechanical tension – unlike muscle damage, mechanical tension has been described as an essential/necessary condition for promoting hypertrophy (3). The reference eludes me but I recall someone somewhere mentioning that even without neuromuscular input (ie without “activation” of the muscle) the application of tension could induce some hypertrophy.

“Tension” entails the application of force and stretch to the muscle, and continual overload in this manner stimulates growth. However, tension itself is not SUFFICIENT, which is where some of the confusion arises. Schoenfield references (in the paper above) lower volume routines with high tension resulting in neural gains only, something which anecdotally rings pretty true if you consider the general lack of impressive hypertrophy experienced by people who follow very low volume strength routines long term as compared with bro-esque training.

So there is a volume element to tension that is important to satisfy for promoting growth. Brian Minor wrote an excellent article (5) about whether load progression is necessary for hypertrophy that writes a lot of stuff that I would otherwise have to write here, so I suggest reading it.

While I’m on tension, and while Brian has hopefully just given you an idea of how volume is a convenient proxy for tension demands, and while you’re maybe even considering that tension-producing demands on an fiber by fiber level drive adaptation, I’ll just write quickly about eccentric loading and hypertrophy.

Concentric and eccentric contractions in isolation can produce similar hypertrophic responses (6), but eccentric contractions induce a greater acute elevation in MPS, presumably because they are more damaging, as the force produced per cross-bridge (the actual little bit that helps muscles contract) is greater while the structure lengthens under load leading to more structural stress.

Combining concentric and eccentric contractions is more effective for growth than either in isolation. There’s probably a few reasons eccentrics help with hypertrophy. Firstly (most obviously), working the muscle in both directions as opposed to one means more actual volume/tension stimulus for a given number of sets and reps.

Secondly, maybe, because they induce greater muscle damage, thirdly, also maybe, because of the fewer cross-bridges formed tension is concentrated further. Finally (and I say this most speculatively of all), there’s some controversy about whether eccentric contractions result in the recruitment of motor units following the size principle in the same way as concentric and isometric contractions, with some evidence that large motor units are preferentially recruited and some conflicting (16, 17, 18).

That could go some way to explaining why type 2 fibers have been observed to be more prone to hypertrophy (they’re already doing the concentric work with moderate to high loads, are (maybe) disproportionately exposed to tension on eccentrics and therefore exposed to more tension overall, and suffer more metabolic disturbance due to their lesser aerobic capacity (and also the greater heat stress associated with eccentric work).

Anyway, that latter portion especially could all be wrong and I’m not sure it matters in practice.

What is definitely interesting is that, given our eccentric strength exceeds our concentric strength, one might expect that eccentric overload training is a viable strategy for muscle growth (because you could expose fibers to greater overall tension and/or damage) but the jury is out on whether that practice has much value (7).

So that’s all fluff but one of the MOST important points of this whole article that I want to make is that while there is a volume component to providing tension stimuli, volume itself isn’t the stimulus, merely a proxy, and that progressive tension overload is what actually makes you bigger long term. Let that sink in.

3 – Metabolic stress – unlike tension, which is essential for growth, metabolic stress appears to play an additive role (3). Schoenfield has mentioned a tension threshold, past which point metabolic stress can play an additive effect by augmenting growth signalling through hormonal responses, cellular swelling, free radical production and upregulating transcription factors.

That’s all a little beyond me tbh but conceptually considering an initial growth threshold of sufficient tension with an additive effect of metabolic stress is helpful to explain why multiple set routines are preferable to single set routines and so on, even if the latter can result in growth.

Importantly, though (this might be the SECOND MOST important point of this article) there is probably an upper limit to the additive benefit of additional per-session metabolic stress, which I will cover more when I start (finally) talking about volume, and even though acute muscle protein synthesis scales with volume, I’m wary of confusing high per-session increases that might be attributable to muscle damage for productive. Plus, from a practical standpoint (as I’ll get to), I’m not sure it matters.

What about training

So that’s a conceptual framework for how training stimuli induce hypertrophy and it only took like 2500 words. Essentially, induce sufficient tension and you induce growth, there is some additive benefit to inducing sufficient metabolic stress.

Quickly – part of the confusion around “time under tension” probably arose from some misconceptions to do with both tension and metabolic stress. Whilst there is a volume component to tension, the reduction in work associated with arbitrarily slowing concentric contractions is likely counterproductive. Slowing eccentric contractions is likely beneficial to some degree, as there are force-producing demands associated with resisting the acceleration of the weight under gravity (ie the creation of tension). However, using explosive concentrics leads to higher motor unit recruitment, and the use of the stretch shortening cycle (a “bounce” at reversal from lowering) can increase load lifted and total work done. Given that the metabolic demands of an activity are directly related to external work done, this can lead to increase in metabolic stress also. Which is all to say, in as far as lifting tempo matters, lifting the concentric fairly explosively and controlling the eccentric without going all “superslow” is probably best.

Given that creating an effective tension stimulus involves performing dynamic contractions over long ranges of motions with reasonable volumes you’re likely to suffer some muscle damage on the way although I consider that largely incidental, and said muscle damage might help a bit.

Where this gets more hazy is in the way people conceptualise volume and frequency.


There is a dose response relationship between volume and hypertrophy (8) with higher volume routines leading to better gains. As I said earlier, multi-set routines are better than single set routines (9, 10) too.

Given that Schoenfield didn’t have the data to establish an upper limit for effective volume, many people (understanding that there still must be one) presume that the highest volume that can be performed during a workout will induce the greatest effect. Again, given the relationship between volume and acute MPS and the additive benefit of metabolic stress, this seems on face value reasonable. Those with a little more nuance may consider that for a given distribution of work (ie frequency of training), the highest volume of training possible where performance can be improved session to session is ideal, which may be closer to the truth although I still have some reservations about saying that which I hope to clarify.

Firstly, there is evidence that on a session by session basis, moderate volumes might be more effective for inducing hypertrophy (10).  Performing 30-60 reps/session, (with more advanced folk likely needing more than that), has been observed to lead the greatest rate of growth. In the review above (by Wernbohm) that corresponded to 4-6 sets per training day per bodypart.

I’ll note the (important) caveat that the majority of people in Wernbohm’s review were training 2-3x/wk. Given that total volume DOES take precedence, you could just as easily argue that if you multiplied 30-60 reps by 2-3 you’d get a ballpark appropriately weekly volume target. I think that’s a little simplistic, as the amount of work you do per session influences its recovery demands, but it’s at least a thought.

On a similar note, those in the sample training with the highest per-session volumes got close to equal gains to those in the more moderate range that was deemed optimal. If (and without taking a deep dive into the actual source material, I can’t confirm this) those in the higher volume brackets trained less frequently, then the fairly negligible difference would be simply attributable to them each performing similar weekly volumes.

Another dissatisfying/inconclusive interpretation is made by Schoenfield in his textbook (9) where he states the review found that similar growth was achieved across a wide range of volumes (from approx 20-100 in the case of the quadriceps). I say dissatisfying because given the evidence for improved growth responses from both higher acute volumes and higher weekly volumes, saying a wide range is equally good without addressing the contextual factors that would make it so seems incomplete. But that is sort of what I’m trying to do this article.

A review by Krieger (cited in the textbook above) also observed a benefit to multi-set routines up to 6 sets/d. The Schoenfield review looking at weekly set volumes came about partly to extend upon this, and given the inability to draw an upper limit of effectiveness in the latter it’s hard to say what a daily upper limit would be either.

However, a relatively recent study comparing a modified German Volume Training routine to a lower volume routine (which was still high volume – the former performed 31 sets/d, the latter 21 sets/d across multiple bodyparts) (20) found that there was no additive benefit to the higher volume routine. In fact, greater gains in trunk mass and a trend for improved arm gains was observed in the lower volume routine. The lower volume routine was also able to progress loads at a greater rate over the intervention and trainees in that group experienced greater strength gains at re-testing. Whilst the researchers were hesitant to conclude that the lower volume routine was BETTER, per-se, the ability to actually progress under those conditions is very important – consider that later in my article as I talk about overload, and it does lend support to the notion that there is an upper limit to the adaptive response to high per-session and weekly volumes, even before performance begins to regress as a result.

Which is all confusing. What I think is closest to the truth is that there is a maximum per session dose that will cause you to grow, above which you would yield no benefit (+/- regress). However, there is likely a dose below that point where you get the maximum marginal benefit per unit of volume which I suspect is in the moderate range. Too low and, while you might induce plenty of tension, you’d fail to get all the additive benefits of metabolic stress.

So what other reasons would we have to aim for more moderate per-session volumes?

Firstly, whilst higher volumes increase the time-course of MPS (with 3 sets having MPS elevated above baseline at 29hrs where 1 set didn’t – 9), MPS appears to run its course within 2-3 days irrespective and past a point greater elevations from higher per-session volumes could be simply a case of confusing damage-induced protein synthesis with actual growth. So whilst you may grow a little more and for a little longer from high per-session doses, you don’t grow forever. Further, given that damage does, as I’ve stated, impair your ability to train subsequently, too much is probably not a good thing from a practical standpoint.
Secondly, neuromuscular factors limit how much volume you can do per session, where past a point the quality of work performed decreases, making a case for distributing volume as improvements in total load lifted/tension produced may lead to greater gains.


So whilst volume is a good proxy of tension demands and therefore stimulus to adapt, there must be something to be said for distributing volume across time. There does appear to be an advantage to higher frequencies when training for hypertrophy (11) although the benefit is small when volume is equated – note that this is as you’d expect. Another recent paper compared doing 2 sessions of 8 sets for a given bodypart with 1 session of 16 sets, finding a small advantage to the former in terms of muscle growth (19).

So, again, it would seem for a given amount of volume that distributing it across the week is beneficial to a degree. Lower per-session volume reduces neuromuscular fatigue, improving total load lifted. Given that MPS runs its course in 2-3 days, one could say that training frequently keeps MPS elevated (and therefore keeps you growing) and if you don’t consider muscle damage helpful, then distributing volume reduces the amount sustained and that could be a plus too. But again, the simplism of “train as frequently as you can” means that it misses the point.

Firstly, as I keep saying, there is a threshold of tension stimulus required to stimulate growth (this is probably pretty low, and feasibly training very frequently with low volumes satisfies this) but there is also additive benefit from metabolic fatigue that entails training with higher volumes. Given that moderate per-session volumes were favoured by Wernbohm, presumably 30-60 reps, or 4-6 sets (more if you’re advanced) are required to get a near-maximal per-session benefit.

Secondly, maxing out your per-session benefit reduces your capacity for force production for longer than MPS remains elevated. Performing 8 sets of bench press resulted in decreased force production capabilities at the 3 day mark post-session (12). Given that the ability to produce high forces (tension) is necessary for inducing hypertrophy, the tradeoff of training at a level of per-session volume that gives the maximum marginal growth benefit is probably that you can only do it every 3-5 days.

Just quickly, in contradiction to the above, Wernbohm actually observed the fastest size gains in the highest frequency programs included in his review, however these routines were very unusual – eg training 12x/wk for 2 weeks with 20% 1RM – not particularly applicable to most. I later mention possibly periodising frequency as a viable strategy, but I consider that a short-term approach and not necessarily your bread and butter.

Schoenfield’s meta analysis on frequency was not able to determine whether training 3 times/wk was superior to 2, but he has stated elsewhere that around 2x/wk appears to be the sweet spot.

In short, my suspicion is that frequency is not that important in and of itself, but rather that optimal frequency exists as a function of the recovery time required between sessions that are maximally/optimally stimulating, and that that just so happens to be 2-3x/wk.

That would also help explain why many very advanced lifters might gravitate towards brosplits or having a hard and easy day for a given bodypart per week, as the recovery time between providing a sufficiently large tension/metabolic stimulus to induce adaptation increases as their threshold for adaptation does, so that only one session might fulfil the full tension and metabolic demands for max stimulus and the other just provide enough tension stimulus to keep things ticking over.

Again, if you’ve skimmed Brian’s article you might say “if force production capabilities are decreased, you can just do a heavy and a light day and the extra volume will compensate for the decreased force you produce in the latter session” but the issue with that is that if your “lower stress” session imposes similar recovery demands as your harder/higher force session, you still won’t be recovered in time to perform at a high level again when you need to. For the sake of completeness – while volume is a big contributor to muscle damage, it is ALSO true that the amount of volume you can tolerate before overreaching is lower at higher intensities (10) so there is probably a degree of middle ground to find in having light top-off days, but not likely so much that you can make both days high stress and that’s the point.

So bringing things ALL the way back to why I have a subtle semantic problem with saying “do as much volume as you can recover from” – it’s close to true, but I’d rather think “people need to apply sufficient training stress to induce adaptation, recover from it, and then reapply that stress” with volume per-session reflecting how much stress is actually required to provoke a response and frequency reflecting the recovery demands of that work. When you’re advanced that might mean your hardest bouts are separated by lighter sessions to allow sufficient recovery. By extension that process gives you a volume per week or month that is about as much as is required.

For advanced lifters, the difference between the above and doing as much as you can recover from is probably very little, but if you’re a beginner your recovery abilities are quite high as compared to the amount of work required to actually induce an adaptation, such that the two aren’t the same at all.

To sum up ALL of volume and frequency (because I’ve contradicted myself enough to be confusing) – total volume is much more important than how you distribute it. The literature seems clear on that. However, on a per-session basis the optimal volume is likely moderate – too low may not be maximally (or sufficiently) stimulating, whilst too high may not yield as much marginal return. Given the recovery demands of moderate volume sessions, you can probably sustain sufficiently overloading sessions 2 (maybe 3) times per week, and once you get big/strong enough to require more work to actually induce much change, it might be less or require some undulation. That said, if you can sustain moderate volumes twice a week that is better than doing it once, and if you are progressing with a moderate volume routine, doing more may be beneficial, but only if you can do more and continue to progress long-term. Which is to say, high volume routines are better probably because they drive adaptation, but chasing higher volumes probably isn’t better past a point if you’re already adapting.
Because there are tradeoffs to both very high frequencies and very low frequencies, neither is likely optimal. However, both are workable, and provided your training is structured such that you can progress, it is probably OK to go with your preference.


Now I really hope you’ve read Brian’s article because it says a lot of brilliant stuff I won’t. There needs to be an increased physiological stress imposed on muscle to tissues over time to continue to induce adaptation, certainly.
However, it is probably more useful to think of overload as an event, rather than a metric. That is to say, a session is overloading if it exceeds the stress (or impulse – to use Brian’s term) threshold to cause adaptation. This threshold creeps up over time, hence the observation that over training careers, people need more work to induce adaptation.

This DOESN’T mean that every session you do must represent a lifetime personal best to be stimulating. Just to hammer out the point – if you consider overload as a metric whereby you have to set some type of tonnage/load PB to achieve anything, you’d expect to see no benefit from submaximal training, yet routinely people DO get stronger training with weights and volumes that they have used before, before setting PBs for the final couple of weeks of a cycle. What’s more, plenty of people bash their heads against the same attempted PB that they keep failing at without making progress.

In the same vein, most people would agree that if you benched a new PB of 100kg for 5 one day and then in your next session did the same thing, provided the effort was high enough that would still induce some stimulus to adapt. However, with enough repeated sessions that stimulus would eventually become insufficient to provoke any change.

So, more accurately, progressive overload is both an indication of the efficacy of prior training and a requirement for ongoing development.

Because it was raised by my colleague, Luke Tulloch, when he read over this article for me – it may be that your best indicators of growth short-term aren’t even the big multijoint exercises. The more potential for neuromuscular adaptations to improve performance (think of technical improvements, antagonist inhibition, improved bracing etc all contributing to increase squat performance) the less you can say improvements arose specifically as a result of added force-producing architecture (muscle). I believe that idea was stated by James Krieger, who has coauthored a couple of big meta analyses in this field, so credit to him.
There’s some indirect support for the relative unimportance of frequency when total volume is controlled that arises from that notion, in that in a recent meta analysis on frequency and strength gains no benefit was found for higher frequency in single joint exercises (13) – if neural factors aren’t driving performance increases in these exercises to the same degree then whichever routine lead to greater hypertrophic gains should win out, yet no advantage was observed. Food for thought.

I still tend to use big basic movements as my benchmarks for progression, because whilst in the short-term their improvement may be confounded by neuromuscular factors to a greater degree, over the long term significant gains will still be largely muscularly driven. The reference eludes me, but consider that among elite powerlifters muscle mass is a predictor of performance in each discipline, and that strength and muscular improvements are closely tied in the longer-term.

I ALSO need to insert a paragraph somewhere talking about why some people also benefit from reducing volume, despite the broadly correct idea that as training careers progress, more volume is required to continue to advance.
The first is the issue of recoverability – if you are jamming in heaps of volume and not actually progressing in any of it, your training sucks (I reiterate that next). Plenty of people adopt an “outwork” mentality or aim to do more work than they’ve ever done before, which is admirable in spirit, but appropriately titrating your training stress to allow you to keep progressing is much more effective.
Again – if you took overload to literally mean inducing more stress than you’ve ever induced before, reducing from too much stress to the right amount wouldn’t be helpful, but we know that not to be the case, even just from anecdote.
The second is in the case of very advanced athletes for whom the recovery demands of truly overloading sessions are super high. I already briefly touched on my speculative thinking for why advanced lifters might move towards bro-splits or heavy/light styles of training. It’s not necessarily because on a per-session basis that structure allows them to grow the most, but rather that to structure their training to allow them to do hard enough sessions to actually induce much change also means a decent proportion of less stressful training – ie, to do much growing at all they might have to grow a little less often, and for each set and rep they perform the actual stress induced is much higher (anecdotally, super strong/big powerlifters often do much less volume than their smaller counterparts, too).

So that’s consistent with my statement that overload is more of an event than a metric, whereby you might just not be able to sustain as frequent overloading events where the requisite stress is super high. Where I’m a little more uncertain is at what point consolidating volume would yield more value than continuing to spread it to cram more in, but the broad thrust of this section has been that the proof of effectiveness of prior overload is in your ability to progress and so I’d stick with that – if you can’t sustainably progress with 2 equally hard sessions per week due to excess fatigue, making one slightly easier might be a logical step and so on.

This is where I combine my previous two MOST IMPORTANT points of the article into one SUPER POINT – if your hypertrophy training never results in you being able to demonstrably induce greater stress over time, it probably sucks. “Optimising” volume/frequency without the direct outcome being an improvement in that metric probably means nothing. You need to train at volumes that induce enough stress to elicit an adaptation, distributed such that you can recover from it, and do it again chronically. The proof is actually in the pudding.

Now, while Brian (correctly, I think) points out that load progression isn’t necessary for hypertrophy, your best metric for training success in this regard is repeated set rep efforts (ie, your best 3×10, or 4×8 etc), where increases are a good indication of growth.

This means hypertrophy training benefits from STRATEGIC PLANNING to actually allow you to lift greater loads long-term. This entails both microcycle-level manipulation (think undulation of volume/intensity between sessions in the week, appropriate volume dosing on given training sessions), mesocycle-level (volume/intensity progression plans over the course of a given block/blocks) and, more tentatively, macrocycle level manipulations.

Given that I’ve already sort of covered microcycle concerns (total volume, volume distribution and frequency) in some respects, and given that I think mesocycle-level concerns are where people most royally fuck it, I’ll talk about that.

If we grant me that overload merely means we exceed the adaptation threshold, and then that gives you some room to start your progression plans submaximally.

So, given that our stimulus is basically an intersection of force production demands and time (ie volume), we need to structure a way to increase it over the mesocycle.

However, because I’ve already said I think there’s probably a rough optimal per-session volume (with an associated recovery cost that basically dictates a per-week volume), we want to spend most of our time with training stress around the maximally beneficial range.

What this means in practice is that we should start somewhere in the ballpark correct volume range (with the small exception of intro weeks – more on this soon) and increase load lifted or reps performed for a number of weeks, deload (more on this soon too) and then repeat, aiming to go further again in each subsequent cycle.

Load lifted and failure proximity are two primary determinants of motor unit recruitment. Increasing the load lifted (duh) for a given number of repetitions naturally takes care of failure proximity, and increasing the number of reps lifted with a given load does likewise. Over the course of training cycles, by increasing loads/volumes we naturally introduce more stimulus, but by starting submaximally we allow greater recovery (as training to failure is disproportionately more fatiguing and may reduce overall volume tolerance 13) and build in some room to progress. Given that it’s been shown that being even 5 reps short of failure with moderate loads can lead to full motor unit recruitment (same citation), submaximal training can still expose all fibers to tension. It’s actually very hard to determine reps in reserve accurately, particularly when performing more than a few repetitions, but one useful method to ascertain whether your sets are difficult enough is whether lifting speed begins to involuntarily slow (again – credit Luke Tulloch).

As you train closer to failure or lift heavier loads the larger/faster twitch motor units will be recruited earlier in the set. Theoretically that could mean the hypertrophic stimulus for those specific fibers would increase a little even if total volume reduced, although I’m less sure of that (as while you may perform only 8 reps in a set, whereas previously you performed 12, if the highest threshold motor units were active for 4 of your 8 reps and only 2 of your 12 their tension exposure would be higher, but yeah). For that reason I don’t think you need to get too fussed if your tonnage drops a tiny bit within a 4 week block if you advance loads and reps fall a little bit to accommodate it, because as I’ve sort of said, if fibers are being exposed to sufficient tension for a sufficient duration to cause them to adapt, then that is all well and good.

Anyway, this section was meant to be about doing things in practice. So in practice what this means is that you ought to be reviewing your current best sets for a given lift and rep range, working backwards a number of weeks (3-4 is a good start) and planning a progression to surpass them. That may mean starting a cycle with 4 or so reps in reserve, and that’s fine.

I’ll quickly cover intro weeks and deloads for completeness, then give an example for progression.
WAY back in the muscle damage part I mentioned that the repeated bout effect might even be a good thing, in that it entails less damage sustained from a given bout of exercise. Given that muscle damage lengthens recovery time and reduces force production capabilities, being able to sustain higher volumes for less damage is likely good, as it means greater exposure to tension/metabolic stimuli for a given amount of recovery.

Most mesocycles I think it is beneficial to have an introductory week with slightly lower intensity and volume to acclimate a little bit to the stimulus. Jumping straight into full volume runs the risk of reducing the degree to which you can advance due to excessive muscle damage and disproportionate fatigue, and this is particularly true when you introduce novel exercises or loading schemes.

A well-structured intro week might be enough stimulus to introduce adaptation, but even if it doesn’t make you much bigger/stronger in and of itself, if it sets a higher ceiling for advancement across your block that is still beneficial.

A downside of training with relatively higher volumes is that doing so hastens the onset of overreaching. Having deload periods is necessary to facilitate recovery and allow ongoing periods of hard training, and whilst plenty of the hypertrophy-obsessed crowd hate the idea of actually resting a bit to get better, it is important to remember that long-term your results are only going to reflect how much you can actually improve your gym performance. There is also at least some theoretical basis to accumulated fatigue impairing growth signaling (9) so there’s competing prerogatives of training hard enough and long enough to make a difference and including enough recuperation to continue to make that possible and maximise your return on investment.
For less advanced trainees deloads are likely less necessary, as the volumes required to grow are lower. However for more advanced athletes where higher volumes are a necessity to get anywhere, and for whom sufficient stimulus on a per-session basis begins to bump up nearer their recovery ceiling, deloading is just a fact of life.

So what could that look like? Here’s an example

I deliberately was vague in exercise selection terms because that’s not as important. What is important is this
– Week 1 is an intro week. You would aim to find a load to hit the target reps with approximately 3-4 reps in reserve.
– Weeks 2-4 are loading weeks (this could easily be weeks 2-6 or similar, depending on how you progress) where you would aim to increase load by the smallest available increment on a weekly basis provided you could fall within the target repetition range. In the final week you would aim for 1 rep in reserve or failure on your last set of most exercises.
– The “deload” roughly halves the volume of exercise performed prior.
– This is also just one day of a program, and presumably you’d perform more upper body work on another day of the week

If you were filling those exercise slots with exercises you already had a baseline of performance for (eg, bench press, where your best performance is 100kg for 3×8) it might look like this (with week 4 pushing for PRs of at least 3×8 – the numbers are meant really to be figurative in this instance, so don’t be too hung up on them

So that’s mesocycle structure. Others (Mike Israetel of Renaissance Periodisation, for one) would probably advocate for something similar although as far as I’m aware his preference would be for adding sets over the cycle and not letting reps fall. I think I’ve made my case for why I’m not convinced the former is necessary and less worried by the latter, but given the otherwise enormous similarities in what we’re saying it’d be remiss not to cite him.

Another program that works on similar-ish principles is Lyle McDonald’s Generic Bulking Routine which has people start at approximately 90% of their best efforts for a given rep range (+/- a small volume reduction) and then ramp up over a couple of weeks before setting PBs for a number of weeks, resetting and starting again.

For structure on the macrocycle level, I’m not going to go into as MUCH depth. I think what I wrote above should represent the bulk of what your training would look like, with refinement over time in terms of the amount of volume you use. A few overarching considerations, though

– It may be worth periodising frequency. Although I tend to believe, as I’ve said, that 2-3x/wk frequency for most bodyparts is probably best, there is the tentative evidence for short term benefits of very high frequency in Wernbohm, and Schoenfield has mentioned that this is a viable strategy (9) to take advantage of lower neuromuscular fatigue and handling higher total loads. Anecdotally, plenty of people suggest upping the frequency of training lagging bodyparts as a means of getting them to grow more, and I suppose this gives some small support to that as a viable short-term strategy, although long-term I suspect specialization cycles might be required.
In very high frequency periods it might be worth using more compound exercises as improved motor learning with higher frequency could improve total load lifted in these exercises more (13) and so they probably tack in nicely with periodic strength phases, especially as higher frequency training with lower per-session volumes will provide a tension stimulus with significantly less metabolic stress. The authors of the prior paper also cite a greater timecourse for recovery in single-joint vs multi-joint exercises, as biceps peak torque remained suppressed for longer in a curl than a row in a study (14)  however I’m skeptical as to whether that is the case, given that it could just as easily be that that was because the row doesn’t primarily train the biceps to the same degree.

In a similarly tentative vein, it might also be worth having periods of lower frequency, bro-split style training to take advantage (if there is any) of muscle damage from higher session volumes. Also, high frequency training more rapidly induces overreaching (9) so at the very least it might present a recoverability benefit. But beyond that I’m not very convinced as I think sufficient volume, range of motion and exercise novelty would probably induce enough muscle damage to keep people progressing, if it’s even necessary.

I don’t think you should change exercises very often. Early strength gains in an exercise likely represent neural adaptation, at least for the most part, and so changing too often runs the risk of never actually accruing new lean tissue to fuel performance increases. Furthermore, new exercises are more damaging/disrupting, so conserving the ones you choose for a while means the benefits of the repeated bout effect are maintained to a greater degree.

That said, you do want to change exercises here and there, maybe every 8-12 weeks or more, as different lines of pull may stress different aspects of the muscle fiber. There is also some evidence for regional hypertrophy depending on joint action and even type of contraction (15, 7) meaning that for full development some variation is probably a good idea. Given that stress isn’t necessarily evenly distributed across a muscle with a given exercise, rotating exercises here and there can also reduce overuse injuries.
Because I like to structure training long-term to facilitate improvements in load lifted, I tend to structure my variations over macrocycles to do the same (eg progressing from a high bar squat with a slow descent to a low bar squat over a number of blocks), however that is a very powerlifting-centric mindset and the increases in load lifted between two similar exercises don’t necessarily correspond perfectly to increases in actual tension demands on the muscle. For pure bodybuilding, it is probably better to consider joint angles where peak stress occurs/lines of pull and emphasise one or two at a given time, then rotate them out after a few cycles. Within that you can still progress exercises to facilitate load increase (eg incline DB press to incline barbell bench) but with less of a premium on such.

– Metabolic stress-focused phases are increasingly popular, but I’m not entirely convinced of their utility, given that tension is an essential requirement for growth and metabolic stress is merely additive. However, metabolic stress phases CAN satisfy tension requirements to grow and low load training CAN work, so they may have their place in a phasic plan.
In a similar sense, strength phases are probably not the best hypertrophic stimulus, even if they do provide sufficient tension to make you grow. However, by improving abilities to lift heavy loads and reducing overall fatigue, they may still be of use periodically.

So a macrocycle structure might look like

Block 1
Similar to my example block above

Block 2
Similar to my example block above, but with 1 more set across the board if recovery allowed it, or simply seeking to surpass the performance benchmarks of block 1.

Block 3
All rep targets reduce by 2-3, exercises change to accommodate slightly heavier loads. Any further alterations to set number necessary.

Followed by either a short strength phase +/- a metabolite phase and then restarting the process.

The take- homes

So I’ve taken >8000 words to say it so far, but the take-home message is that hypertrophy training isn’t THAT complicated, really. It just benefits from more strategic planning than people give it.

The conceptual research on the topic has given us the insight to say that higher volumes are helpful, and that training twice a week is better than training once, however beyond that the proof of efficacy of training is in your ability to adapt to it. Whilst there are theoretical “optimals” for each training variable, they should be used more as signposts to guide the construction of a program that allows you to progress the stimulus applied.

That means that the vast majority of the time you should just
– train everything twice a week, or thereabouts
– do relatively moderate volumes each session. The evidence suggests that >10 sets/wk likely maximizes growth, and my hunch is that something like 4-10 sets per bodypart per session is probably good for most people too, so those numbers jive. (Wernbohm found the best growth at 4-6 sets a week, I’m just making some accommodation for more advanced people and the rough timecourse of recovery from doing 8 sets in another study I cited, plus the fact that most of Wernbohm’s sample trained to or close to failure and not everybody does that in practice).
– use the same exercises for 8-12 weeks at a time, or more.
– progress from submaximal efforts to PBs across each cycle, and plan your progressions so that you actually get better. Impose an overarching structure on your training so that cycle to cycle to you can progress, too.
– It’s not just OK to have some sessions/weeks that don’t grind you into the dirt, it’s probably actually beneficial. Likewise, you’ll probably have to occasionally deload.

And if ALL of the efforts to optimize your training fail, if you are still lifting more weight for the same or more reps over multiple sets, you’re doing fine.

That’s because – at the end of the day, it’s not actually doing progressively more volume that drives adaptation, or having your frequency high enough to constantly stimulate MPS, it’s just imposing enough training stress to induce adaptation, recovering enough to do it again, and rinsing and repeating. All the training variables are is a means to structure training to ensure that on an ongoing basis.

If you LIKED this article, join my mailing list – it’s FREE and I can send you stuff directly to your inbox. You can also follow me on Instagram, where I frequently post training and diet analysis and advice on my stories.

I also have a podcast, Weakly Weights, available on iTunes and Podbean, where I discuss training for powerlifting in-depth. You can join our mailing list, too, for free sample programs when we discuss them.

Finally, I have now made home workout templates available for sale. They’re designed to be used with any combination of bodyweight, bands, kettlebells/dumbbells and weighted backpacks. They individualise volume prescriptions and offer feedback on when to progress your training to get the most out of limited equipment. You can find them here.

Addendum 2019 – Further reading

This post by Menno Henselmans discusses the potential for a maximum productive volume per session and its interplay with total weekly volume/frequency. Referencing two studies that came out after I wrote this article, and a third that I discussed, he comes to quite similar conclusions to my own.

This study was released and caused a massive internet shitfight (for a small portion, see here), with some commentators questioning the researchers’ methods and their interpretations of their data. There was a reported trend towards greater growth with volumes far higher than had been previously observed.
A large number of caveats need to be applied before concluding that 45ish sets/wk to failure on a muscle group is optimal, and (similarly to this study, which was released shortly afterwards), some commentators have also argued that what the paper demonstrates is that higher volumes of work may be necessary if the quality of the sets are lower (ie if the loads are submaximal, rest is insufficient, and “failure” is not true failure).
It is also not clear whether short-term very high volume phases are beneficial as compared to more moderate volumes, or whether chronically high volumes yield better adaptations than more moderate volumes. Purely anecdotally, many bodybuilders have pursued short-term specialisation phases with much higher volumes for 1-2 muscle groups at a time with success. It may be that short-term research demonstrating a small marginal benefit to very high volumes supports the use of short-term high volume approaches within an overall more moderate backdrop.

Lyle McDonald (who, it should be stated, has made a personal mission of “debunking” this paper) wrote a 3-part review grouping studies on training volume, with his estimation of the optimal weekly volumes falling in line with the 10-20ish that most other names in the field have put forward. I’ve linked part 3 here, although parts one and two are worth reading also, bearing in mind that a large part of his analysis is directed towards discrediting the paper mentioned above.

Lyle also wrote an absolutely excellent series on muscular tension and how it applies to training/load progression, which should be read in it’s entirety. If you don’t have time, part 3 addresses most of the training-related questions.
Part 1
Part 2
Part 3


 1 – The development of skeletal muscle hypertrophy through resistance training: the role of muscle damage and muscle protein synthesis.

Eur J Appl Phys 2018

2 – Regenerated rat skeletal muscle after periodic contusions

Braz J Med Biol Res. 2001


J Strength Cond Res. 2010

4 – Does exercise-induced muscle damage play a role in skeletal muscle hypertrophy?

J Strength Cond Res. 2012

5- http://myojournal.com/load-progression-hypertrophy/

6 – Early structural remodeling and deuterium oxide-derived protein metabolic responses to eccentric and concentric loading in human skeletal muscle

Physiol Rep. 2015

7 – Eccentric Overload Training: A Viable Strategy to Enhance Muscle Hypertrophy?

Strength and Conditioning Journal – 2017

8 – Dose-response relationship between weekly resistance training volume and increases in muscle mass: A systematic review and meta- analysis

Journal of Sports Sciences – 2016

9 – Science and Development of Muscle Hypertrophy – Schoenfield

10 – The Influence of Frequency, Intensity, Volume and Mode of Strength Training on Whole Muscle Cross-Sectional Area in Humans

Sports Med 2007

11 – Effects of Resistance Training Frequency on Measures of Muscle Hypertrophy: A Systematic Review and Meta-Analysis

Sports Med 2016

12 – “Dissociated Time Course between Peak Torque and Total Work Recovery Following Bench Press Training in Resistance Trained Men.”

Physiol Behav. 201

13 – Effect of Resistance Training Frequency on Gains in Muscular Strength: A Systematic Review and Meta-Analysis

Sports Med 2018-

14 – Dissociated Time Course of Muscle Damage Recovery Between Single- and Multi-Joint Exercises in Highly Resistance-Trained Men

J Strength Cond Res. 2015

15 – http://www.lookgreatnaked.com/blog/can-you-target-different-aspects-of-the-hamstrings/

16 – Eccentric contractions require unique activation strategies by the nervous system

J App Physiology – 1996

17 – Can fast-twitch muscle fibres be selectively recruited during lengthening contractions? Review and applications to sport movements

Sports Biomech. 2008

18 – Differences in activation patterns between eccentric and concentric quadriceps contractions

Journal of Sports Sciences – 2002

19 – Effect of Resistance Training Frequency on Neuromuscular Performance and Muscle Morphology after Eight Weeks in Trained Men

J Strength Cond Res. 2018


J Strength Cond Res. 2017


Although I don’t cite them directly here, a lot of my thinking is informed by things I have heard/read/watched featuring Lyle Mcdonald, Eric Helms, Mike Israetel and Greg Nuckols, and so their respective writings are probably worth following up on.

Leave a reply

Your email address will not be published.