How many sets per week per muscle, and how to count them

Twelve to 20 hard sets per muscle a week is where most trained lifters should aim, and the low end is the place to start. That band comes from a meta-analysis of trained men in which 12 to 20 weekly sets matched more than 20 for the quads and the biceps [1]. To set how many sets per week per muscle you program, begin near 12, add sets while your logged reps and RPE hold, and stop adding when they slip.

The number is the easy part. The part most plans get wrong is the counting, because a set of bench press does not only train the chest. Below is where the range comes from, how to count a set that trains two muscles, and a worked week with the credits a logger would give it.

Where does 12 to 20 sets come from?

The first useful answer was a dose-response meta-analysis. Pooling 34 treatment groups from 15 studies, it found that each extra weekly set was linked to a 0.37 per cent larger gain in muscle size [2]. When sets were split into fewer than 5, 5 to 9 and 10 or more per muscle, the higher groups grew more, but that comparison was only a trend and not a firm result [2]. More sets grew more muscle. Nobody could yet say where the curve flattened.

A training study then pushed the top end. Trained men did one, three or five sets per exercise, three sessions a week, for 8 weeks [3]. Strength rose about the same in every group. Muscle size did not: the higher volumes grew the elbow flexors and the thigh more [3]. The low-volume group still got stronger on three 13-minute sessions a week, which is worth knowing on a busy month.

The 12 to 20 band itself is from a later review that sorted trials into under 12, 12 to 20 and over 20 weekly sets [1]. Between the moderate and the high groups there was no difference for the quads or the biceps. The triceps did better on more than 20 [1]. The authors offered 12 to 20 weekly sets per muscle group as a standard recommendation for young, trained men, and that is the honest scope of it: seven studies, mostly men, mostly under 35.

The largest analysis so far, 67 studies and 2,058 participants, agrees on the direction and adds the shape [4]. Muscle size keeps rising as weekly sets rise, with diminishing returns. Strength rises too, but flattens much sooner. If your goal is a bigger squat rather than bigger quads, the extra sets past the middle of the range buy you less.

A set only counts when it is hard

Every one of those ranges counts working sets taken to failure or close to it. A systematic review of 14 trials in lifters with at least a year of training concluded that a count of sets is a fair measure of volume when the sets are near failure and the reps fall between 6 and 20 or more [5]. A set you stopped six reps short of failure is not the set the studies counted.

RPE is how you check. RPE (rate of perceived exertion) is a 1 to 10 rating of how hard the set was, and RIR (reps in reserve) is the same scale read the other way: RPE 8 means about two reps left. A working set a couple of reps from failure fits the definition. A light ramp single does not, which is why warm-ups stay out of the total. We covered that case in full in should warm-up sets count as volume.

Count indirect sets as half

A bench press set is one chest set. It also loads the triceps and the front delts, just less directly. If you count it as zero for the triceps, you underrate how much arm work you already do. If you count it as a full triceps set, you overrate it.

The research now has an answer for this. The 2026 meta-regression classed every set in every study as direct or indirect for the muscle being measured, then tried three ways of counting the indirect ones: as a full set, as half a set, or as nothing [4]. The evidence was strongest for half. The authors concluded that telling direct from indirect sets apart appears essential for predicting what a program will do [4].

Top Sets counts weekly volume the same way. A set credits its primary muscle with 1 and each secondary muscle with 0.5. One working set of barbell bench press, logged with chest as the primary and triceps and front delts as secondaries, puts 1 set on chest, 0.5 on triceps and 0.5 on front delts. Warm-ups credit nothing. A drop set or a myo-rep set credits half, so one drop set of dumbbell curls is 0.5 for the biceps and 0.25 for the forearms. The app tracks this as weekly sets per muscle for 14 muscles and shows each in plain English against a target.

That half credit is why small muscles often look covered before you add a single curl. It is also why the front delts can need almost no direct work in a program with a lot of pressing.

A worked week

Here is one upper-body session, done twice a week. The working sets are what a lifter would log: 4 sets of barbell bench press after 2 warm-up sets, 4 sets of seated cable row, 3 sets of pull-ups, 3 sets of cable triceps pushdowns, 2 sets of dumbbell curls plus 1 drop set, and 3 sets of dumbbell lateral raises.

The weekly credits below are the output of the app's volume model on those sets, run through compute.mjs.

Read it from the top. Back sits at 14, inside the band. The biceps reach 12 with only 5 direct sets a week (four curls and two half-credit drop sets), because rows and pull-ups add 7. That is more than the triceps get, even though the session has more pushdowns than curls. The 4 warm-up sets of bench press added nothing.

Chest is the gap. Eight sets is below the 12 the research suggests as a starting point, and the fix is cheap. Add one more set of bench press and one set of barbell incline bench press to each session, and the model puts chest at 12. The same change lifts the triceps from 10 to 12 and the front delts from 4 to 7, with no extra arm work. Side delts at 6 are low too, and lateral raises are the only thing in the week that trains them directly. The rear delts get 4 from rows alone, which is a reasonable floor but not a target.

The table also shows what not to fix. Adding curls to bring the biceps level with the back would push them well past 12 for a small muscle that already gets most of its work for free.

How to set and adjust your targets

Pick a starting number per muscle, then let the log move it. A plan that works for most trained lifters looks like this:

• Start at about 12 weekly sets for the muscles you most want to grow, and near the bottom of the band for the rest.
• Count primaries as 1 and secondaries as 0.5 before you add direct work for small muscles.
• Add sets a little at a time while your reps hold at the same load and RPE.
• Stop adding when reps drop at the same RPE across three sessions (the app's default for flagging a stall), or the same reps start costing a higher RPE.

The third and fourth lines are where a log earns its place. If your bench press reps fall at the same load and RPE across a few sessions, more sets are not the answer that week. That is the signal the research cannot give you, because the ranges are averages across groups and you are one lifter.

Volume also goes up and down on purpose. A training block usually builds volume over its weeks, and a deload cuts it to let fatigue clear. How often to take one is in how often should you deload.

Mistakes that skew the count

Counting every set on the same line is the most common one. A lifter who writes "chest: 20 sets" after 12 sets of bench press and 8 sets of dips is right about chest and silent about the triceps, which just got a block of work nobody planned for them.

Counting easy sets is the second. Twenty sets at RPE 5 is not 20 sets. The research behind the band counted sets near failure [5], so a target of 16 sets is 16 hard sets, not 16 lines in a notebook.

Chasing the top of the band before the bottom has stopped working is the third. More sets keep paying off for size, but each one pays a little less [4]. A lifter who still grows on 12 sets has no reason to run 20, and every set above that still costs recovery.

The last is changing too many numbers at once. If you add sets, raise the load and change the exercise in the same week, the log cannot tell you which change did what. Move volume on its own.

The app shows each muscle's weekly sets against its target on the Progress tab, so the gaps in the table above show up without a spreadsheet. The home page shows how the weekly volume readout looks, and the free lifting tools cover the numbers around it, from an estimated one-rep max to loading the bar.

Questions

Do indirect sets count toward weekly volume?

Yes, at about half. The largest meta-regression on volume compared counting indirect sets as full, half or zero, and the evidence was strongest for half. So a set of bench press is one chest set and roughly half a triceps set. Counting it as zero underrates your arm work, and counting it as a full set overrates it. Top Sets credits secondary muscles at 0.5 for that reason.

Is more than 20 sets per muscle a week too much?

Not for every muscle, but rarely worth starting at. In the review behind the 12 to 20 band, more than 20 weekly sets did no better for the quads or biceps, and better for the triceps. The larger meta-regression shows muscle size keeps rising with sets but flattens. Go past 20 only when your reps still climb at 20 and recovery holds.

How many sets per week to build strength rather than size?

Fewer than for size. In one trial, trained men doing one set per exercise three times a week got about as strong as men doing five sets, while the higher volumes grew more muscle. The largest analysis found that strength flattens at lower volumes than muscle size. If strength is the goal, spend effort on load and practice rather than extra sets.