Optimal Training Volume and Frequency: How Many Sets per Muscle Actually Build Muscle

For most lifters, the optimal training volume for muscle growth is roughly 10-20 hard sets per muscle per week, and 6-10 sets already gets you a big share of the gains. More sets do build more muscle, but every extra set pays back less than the one before. How often you train a muscle matters far less than how many good sets you get through in the week.

That is the short answer. The longer one is more useful, because “10-20 sets” means nothing if you count sets wrong, cram them into one exhausted session, or pad them with half-effort reps.

This guide walks through what the dose-response studies actually measured, where productive volume ends and junk volume begins, how to count sets from compound lifts, and how to build a week you can recover from. It sits alongside our guide to muscle hypertrophy for natural lifters, which covers the bigger picture of how muscle grows.

What is the optimal training volume for muscle growth?

Training volume, in this guide, means hard sets per muscle per week. Not tonnage. Not total reps. Hard sets, meaning sets taken reasonably close to failure (we explain why that unit wins in hard sets vs tonnage).

Here is the working answer, then the evidence behind it:

• Under 5 sets per week: enough to maintain, and enough for real beginners to grow, but usually not enough to push trained muscle much.
• 6-10 sets: a strong starting point. Plenty of trained people grow well here.
• 10-20 sets: the range where most research shows growth beginning to level off.
• 20+ sets: sometimes a little more growth, sometimes nothing extra, always more fatigue.

The honest version: there is no single optimal training volume for everyone. There is a curve, and your job is to find where you sit on it.

What the dose-response research actually shows

The case for “more sets, more muscle” rests on several meta-analyses, and they agree on direction more than on exact numbers.

Krieger (2010) was the first big look at hypertrophy. Pooling 8 studies, multiple sets per exercise produced about 40% larger hypertrophy effect sizes than single sets, in both trained and untrained people (Krieger, 2010). Effect sizes crept up from 2-3 to 4-6 sets per exercise, but the difference between those two was not significant.

Schoenfeld and colleagues (2017) then looked at weekly sets per muscle across 15 studies. Each extra weekly set was associated with roughly 0.37% more muscle growth (Schoenfeld et al., 2017). When they split volume into under 5, 5-9 and 10+ sets per muscle per week, there was only a trend (p = 0.074) favouring the higher categories. That is the paper behind the popular “10+ sets” advice, and it is worth knowing the categorical result was not statistically significant.

Pelland and colleagues (2026) ran the largest analysis yet: 67 studies and 2,058 participants. Gains in both size and strength kept increasing as volume went up, but with diminishing returns, and those diminishing returns kicked in more sharply for strength than for size (Pelland et al., 2026).

Put plainly: the first few sets are worth a lot. Sets 15-20 are worth something. Set 30 is worth very little, and it costs you.

For a deeper walk-through of the curve itself, see the weekly set volume dose-response curve.

Where does 10-20 sets come from, and is more ever better?

This is where the trials disagree, and it is worth seeing why.

Study	Who	Weekly volume compared	What happened to muscle size
Schoenfeld et al., 2019	34 trained men, 8 weeks	1, 3 or 5 sets per exercise, 3 sessions a week	Higher volume grew elbow flexors and thighs more
Heaselgrave et al., 2019	49 trained men, 6 weeks	9, 18 or 27 biceps sets	All grew; no significant difference; 18 sets looked best on effect size
Brigatto et al., 2022	27 trained men, 8 weeks	16, 24 or 32 sets	32 sets beat 16 for thighs and triceps
Enes et al., 2024	31 trained men, 12 weeks	Constant sets vs adding 4 or 6 sets every 2 weeks	No significant difference; possible small benefit that seemed to plateau
Baz-Valle et al., 2022 (review)	Trained men, 7 trials	Under 12, 12-20, over 20 sets	12-20 matched 20+ for quads and biceps; triceps favoured 20+

The case for very high volume. In Schoenfeld’s 2019 trial, muscle growth showed a dose-response, with more growth in the five-set group at several sites, while strength gains were similar across groups (Schoenfeld et al., 2019). Brigatto’s group found 32 weekly sets produced more lower-body and triceps thickness than 16 (Brigatto et al., 2022).

The case for a plateau. Heaselgrave’s biceps trial found no significant difference between 9, 18 and 27 sets, with the 18-set group posting the largest average increase (9.5% versus 4.3% and 5.4%) (Heaselgrave et al., 2019). Enes and colleagues progressively added sets for the legs of experienced lifters (5 years of training on average) and saw clearly better strength, but only a possible small benefit for size that appeared to level off at the higher volumes (Enes et al., 2024). The 2022 review by Baz-Valle concluded that 12-20 weekly sets may be the best standard recommendation for young trained men (Baz-Valle et al., 2022).

Why might they differ? Small samples (often 9-17 people per group), short durations (6-12 weeks), different muscles, and different definitions of a “set.” Ultrasound measurements taken soon after hard training also raised a concern: was the extra “growth” just swelling? A 2026 study in 13 trained men tested this acutely. After sessions of 7, 14 or 21 lower-body sets, muscle thickness returned to baseline within 24 hours, with no sustained swelling, although 21 sets did produce higher effort ratings and worse perceived recovery (Alvarez et al., 2026). That is reassuring for the high-volume studies, but it was a short study of single sessions.

Is there a reason natural lifters might tolerate less? Possibly, though direct comparisons are thin. We cover that debate in natural vs enhanced volume tolerance.

Junk volume: where productive sets stop

“Junk volume” gets thrown around a lot. A useful definition: sets that add fatigue without adding meaningful stimulus. It is not a fixed number. Set 14 might be productive for one lifter and junk for another.

In practice, junk volume usually looks like one of these:

• Sets stopped far from failure. Ten easy reps with a weight you could lift 20 times are not a hard set, no matter what your log says.
• Sets after your performance has crashed. If your fifth set of curls is half the reps of your first, the stimulus per set has dropped, but the fatigue keeps coming.
• Ugly, momentum-fueled reps. Swinging the weight shifts work away from the target muscle.
• Redundant exercises. Three different flat pressing variations back to back, each started already fried.
• Volume you cannot recover from by the next session for that muscle. If you come back weaker, the extra sets probably cost more than they gave.

Here is the part that gets missed. Junk volume is not mainly about the number. A lifter doing 12 genuinely hard sets can be training more productively than one doing 25 sets where half are sloppy. Effort and execution decide whether a set counts. How close to failure each set should go is covered in our load, rep range and RIR guide.

More on spotting it in junk volume explained, and on the old low-volume, high-intensity school in HIT vs high-volume training.

How to count sets: direct sets and fractional volume

You cannot follow a set target if you do not know what counts as a set for which muscle. A bench press works your chest, but your triceps and front delts are clearly working too. Do those count?

The 2026 Pelland meta-regression tested three ways of counting: indirect sets counted as zero, as a full set, or as half a set. The fractional method (half a set) fit the data best (Pelland et al., 2026).

Practical rules for counting:

1. Direct sets count as 1. Bench for chest, squats for quads, rows for upper back.
2. Clear helper muscles count as 0.5. Triceps on presses, biceps on rows and pulldowns, front delts on chest presses, glutes on squats.
3. Only count hard sets. Warm-ups and easy feeder sets do not go in the total.
4. Do not count stabilisers. Your abs work during a squat. That does not make squats an ab session.

This matters most for arms and front delts. Many lifters doing a typical push-pull routine already accumulate several fractional sets for these muscles before a single curl or pushdown. Our deep dive on direct vs indirect fractional set volume works through full weekly examples, and the training volume calculator does the counting for you.

Volume landmarks: MV, MEV, MAV and MRV explained

The volume landmarks idea comes from a coaching framework, popularised by Mike Israetel and colleagues, rather than from a single study. It gives names to points on the dose-response curve:

Landmark	What it means	Rough idea in practice
MV (maintenance volume)	Sets needed to keep the muscle you have	Often low, a handful of hard sets per week
MEV (minimum effective volume)	The least volume that produces noticeable growth	Where a training block might start
MAV (maximum adaptive volume)	The range that gives the best return	Where most productive training sits
MRV (maximum recoverable volume)	The most you can do and still recover	Pushing past this means going backwards

The concept matches the research shape well: low volume maintains, more volume grows, and past some point you cannot recover. The specific numbers you see attached to each muscle online, though, are coaching estimates. They have not been tested in controlled trials muscle by muscle, and they shift with training age, sleep, diet, stress and how hard each set is.

So treat landmarks as a vocabulary, not a prescription. Two of our guides go further: volume landmarks MV, MEV, MAV and MRV and how they change across novice, intermediate and advanced lifters. For maintenance specifically (busy months, exam season, travel), see the science of maintenance volume.

Training frequency: is 2x or 3x a week better?

Short answer: when weekly sets are matched, frequency barely matters for muscle size. It matters a bit more for strength.

The evidence has shifted over a decade, which makes this a good example of how sports science corrects itself.

• 2016: A meta-analysis of 10 studies found higher training frequencies beat lower ones (effect sizes of 0.49 versus 0.30) and concluded that twice a week beat once a week on a volume-equated basis. Whether three times beat twice could not be determined (Schoenfeld et al., 2016).
• 2019: With 25 studies, the same research group found no significant difference between frequencies when volume was equated, including in trained lifters and with direct measures of muscle size. When volume was not equated, higher frequencies had a modest edge, probably because more sessions usually meant more sets (Schoenfeld et al., 2019).
• 2026: Across 67 studies, the effect of frequency on hypertrophy was compatible with being negligible, while strength gains did rise with frequency, again with diminishing returns (Pelland et al., 2026).

Two trials in trained men line up with this. Training muscles 2 versus 3 times a week for 10 weeks produced similar gains in strength and muscle thickness, though effect sizes slightly favoured the twice-a-week split (Lasevicius et al., 2019). Going from 3 to 6 sessions a week with volume and intensity equated did not add strength or fat-free mass over 6 weeks (Colquhoun et al., 2018).

So why does everyone still recommend hitting muscles twice a week? Because frequency is a tool for distributing volume. Sixteen sets of chest in one session is miserable, and the last several are likely done tired and sloppy. Split across two sessions, every set is done fresher. That is the real reason 2x usually wins in the gym, not some magic of protein synthesis timing.

For the full comparison, see training frequency 1x vs 2x vs 3x with equated volume. For how this shapes your weekly layout, our workout splits and session architecture guide and full body vs upper/lower vs PPL volume distribution go deeper.

How many sets per session is too many?

There is no well-established per-session cap. You will often hear “no more than 6-10 hard sets per muscle per session.” Treat that as a coaching rule of thumb, not a research finding.

What we do know points the same way:

• In Heaselgrave’s trial, 9 biceps sets done in one weekly session were enough to grow the muscle (Heaselgrave et al., 2019).
• In Alvarez’s study, 21 lower-body sets in one session produced higher ratings of effort and worse perceived recovery than 7 or 14 sets (Alvarez et al., 2026).
• Rep performance tends to fall across sets as fatigue accumulates, which is the practical sign that the next set is giving less.

A rough rule of thumb, not a tested cutoff: if your reps on a given exercise have dropped by around a third or more from your first working set at the same weight, you are probably in diminishing-returns territory for that muscle today. Move the remaining sets to another day. More detail in per-session set volume caps.

Recovery capacity: why your number is not someone else’s

Two lifters on the same program can respond very differently. Your volume tolerance depends on things outside the gym as much as inside it:

• Sleep. Short or broken sleep is one of the quickest ways to lower the volume you can recover from. Our sleep and recovery guide covers the details.
• Calories. In a deficit, recovery usually gets harder. Most people do better holding or trimming volume while cutting, as covered in volume adjustments during cutting.
• Protein. Hitting a solid daily protein intake supports recovery. For vegetarian lifters in India, that often means planning around dal, paneer, curd, soya chunks and milk rather than hoping roti and rice get you there.
• Life stress and work hours. A 10-hour shift plus a long commute eats into recovery the same way extra sets do.
• Heat. Training in a hot, humid gym in May is harder than the same session in December. Expect sessions to feel tougher and adjust effort rather than forcing extra sets.
• The muscle itself. Some muscles seem to tolerate more sets than others, discussed in muscle-specific volume tolerance.
• Training age. Beginners grow on less. Advanced lifters may need more, but have less room to waste.

There is also a cost to high volume beyond sore muscles, covered in the biological cost of high volume. For building a personal picture of what you tolerate, see individual recovery profiling.

Building your week: practical volume by experience level

These ranges are practical starting points based on the research above, not prescriptions tested as-is.

Lifter	Weekly hard sets per muscle (direct + half of indirect)	Frequency per muscle	Notes
Beginner (under 1 year)	6-10	2-3x	Focus on technique and adding load. More sets rarely needed yet.
Intermediate (1-3 years)	10-16	2x	Most people make good progress here for a long time.
Advanced (3+ years)	12-20, occasionally more for lagging muscles	2-3x	Add volume slowly and only where progress has stalled.
Cutting or high-stress phase	Roughly 6-12	2x	Keep intensity high, trim sets before you trim heavy lifts.
Maintenance (busy period)	A few hard sets	1-2x	Enough to hold muscle until life calms down.

How to adjust from there:

A sample week for an intermediate lifter on an upper/lower split might put chest at about 12 sets (6 per upper day), quads at about 12 sets (6 per lower day), and arms at 4-6 direct sets each on top of the fractional work from presses and rows. If you want a ready-made structure, our free push, pull and legs plans and the leg day plan already balance volume across the week.

Want something more responsive than a fixed plan? Autoregulating set volume shows how to adjust sets on the fly using readiness and rep performance.

Common volume mistakes

• Copying an advanced lifter’s program. Their 25 sets of chest may be what they need after a decade of training. You probably do not.
• Counting every set fully. Logging bench press as a full triceps set makes your arm volume look higher than it effectively is, and makes extra pushdowns seem unnecessary when they might not be.
• Ignoring indirect work entirely. The opposite problem. Adding 10 direct triceps sets on top of heavy pressing can push arms past what they recover from.
• Adding volume to fix poor effort. Covered above. Harder sets first, more sets second.
• Never deloading. Fatigue builds quietly. A planned lighter week every 4-8 weeks, or whenever performance slides, is cheap insurance. See deloads and volume resensitisation and our progressive overload, deloads and plateaus guide.
• Changing everything at once. If you change sets, exercises and frequency in the same week, you will never know what worked.

When to see a doctor

Normal training fatigue clears with a few easier days. Some signs are worth a medical check rather than another deload:

• persistent fatigue, low mood or poor sleep that does not improve after one to two weeks of reduced training
• a resting heart rate that stays unusually high for days
• frequent illness, unexplained weight loss, or, in women, missed periods
• dark, cola-coloured urine or severe muscle pain and swelling after a session, which needs prompt medical attention
• joint pain that is sharp, localised or getting worse rather than better

Overtraining syndrome is uncommon and is diagnosed by ruling other things out, which is a job for a doctor, not a training log. Our article on overtraining syndrome vs non-functional overreaching explains the difference.

The bottom line

The research gives a clear shape and a fuzzy endpoint. More weekly sets build more muscle, with diminishing returns that usually become obvious somewhere between 10 and 20 hard sets per muscle. Above 20, some trials show a little more growth and others show nothing, and fatigue keeps rising either way. Frequency is mostly about spreading those sets so each one is done well.

So: count your sets honestly (direct sets plus half of indirect), make each one hard, start near 10 per muscle, and let your progress over 3-4 weeks tell you whether to add, hold or cut. That is the whole system. If you want the numbers done for you, plug your current program into the training volume calculator and see where each muscle actually lands.