How Often Should You Train Each Muscle Group?

TL;DR: Frequency is not a growth stimulus. It's a scheduling tool for volume. Volume-equated research finds no meaningful hypertrophy difference between training a muscle once, twice, or three times a week. What frequency does buy you is quality: past roughly 11 hard sets in a single session, extra sets stop paying, so higher weekly volumes need more days to land properly. Twice a week per muscle is the right default for almost everyone. Strength is the one outcome that seems to like frequency for its own sake.

---

Ask this question in any gym and you'll get an answer delivered with total confidence and nothing behind it. Bro splits are dead. You have to hit everything twice. Three times if you're natural. The chest needs a week to recover.

The research is much less dramatic than any of that, and much more useful.

The volume-equated answer: frequency barely matters

Schoenfeld, Grgic and Krieger ran the meta-analysis everyone quotes and almost nobody reads carefully (Schoenfeld et al., 2019). Across studies comparing training frequencies, higher frequency looked better. But when they restricted the analysis to studies where weekly volume was matched between groups, the advantage largely evaporated. The higher-frequency groups weren't winning because of frequency. They were winning because more sessions had quietly delivered more total sets.

The most recent and most rigorous look at this comes from Pelland and colleagues, a Bayesian dose-response meta-regression covering 67 studies and 2,058 participants (Pelland et al., 2025). Their finding on hypertrophy was that the frequency slope was compatible with a negligible effect. Weekly volume drove growth. Frequency, once volume was accounted for, did close to nothing on its own.

So if you're doing 14 sets a week for your back, the research says it doesn't much matter whether that's one savage Saturday or two Wednesdays and a Sunday.

Except it does, and here's why.

The per-session ceiling is where frequency earns its keep

Remmert and colleagues took a different cut at the same literature and asked how much a single session can actually absorb (Remmert et al., 2025). Their meta-regression found the relationship between per-session sets and growth is logarithmic, not linear: rapid gains from the first several sets, then a flattening. The point past which extra sets in one workout no longer produced detectable superiority landed around 11 fractional sets per muscle per session.

That number reframes the whole question. Frequency isn't competing with volume. It's the mechanism that lets you spend volume well.

Run the arithmetic. If your quads need 18 weekly sets, one leg day means sets 12 through 18 are being performed on a muscle that has already stopped responding meaningfully in that session, by a lifter whose bar speed died forty minutes ago. Split it into two days of nine and every set lands in the productive zone. Same weekly volume, considerably more of it doing work.

This also explains why low-volume lifters can get away with anything. If you're doing eight sets a week per muscle, one session absorbs all of it comfortably and adding a second day changes nothing. That's exactly what the volume-equated studies keep showing, and most of them use modest volumes.

The practical rule falls out on its own. Your frequency should be set by your weekly volume, not by a philosophy. Up to about 10 sets per muscle per week, once is fine. From 10 to 20, use two. Above 20 you want three, though very few people training naturally need to be up there. Start by working out where your number should sit in the first place: how many sets per week actually builds muscle.

Strength is the exception

Frequency is not a wash for everything. In the same Pelland analysis, the strength result went the other way: the probability that more frequent training improved strength was essentially certain, even with volume held constant.

That makes sense mechanically. Strength is partly a skill. More exposures to the squat pattern means more practice reps, a better groove, better bracing, better load calibration. A lifter squatting three times a week gets 156 rehearsals a year that a once-a-week squatter doesn't. If your goal is a bigger number rather than a bigger muscle, spread the main lifts out. It also helps to know how much of your estimated max is real and how much is arithmetic, because the gap is wider than most lifters assume (how accurate are 1RM calculators). You can run your own numbers with the 1RM calculator.

Picking the split

Once you accept that frequency is a container for volume, split selection gets easy. You're not choosing an ideology, you're choosing how many days you'll realistically train.

Two days a week: full body, both days. Everything gets hit twice. This is the highest-value schedule in lifting for the time invested.

Three days: full body three times, or an upper/lower/full rotation. Still two-plus exposures per muscle.

Four days: upper/lower, twice through. The cleanest four-day structure in existence and hard to beat.

Five or six days: push/pull/legs run twice, or upper/lower plus an accessory day. Only worth it if your weekly volume is genuinely high enough to need the extra sessions. If you're doing six days at 10 sets per muscle per week, you've bought yourself commute time and nothing else. The trade-offs are laid out in push/pull/legs vs upper/lower and full body vs split routines.

One thing worth flagging, because it's the most common way people get this wrong: bumping frequency without cutting per-session volume is not a neutral change. Going from one leg day of 16 sets to two leg days of 16 sets is a 100% volume increase dressed up as a scheduling tweak, and it's a fast route to a stalled squat. Move the sets. Don't duplicate them.

What to actually do

Count your current weekly sets per muscle. If any muscle is getting more than about 11 hard sets in a single session, you have a distribution problem, not a volume problem. Move the overflow to a second day and change nothing else. Give it six weeks and watch whether the last few sets on each muscle stop feeling like an obligation.

Then keep progressively overloading inside that structure, because frequency and volume only set the table. Load and effort still do the eating.

The reason this gets messy in practice is that it requires tracking weekly sets per muscle across a rotating schedule, and almost nobody keeps that up by hand for long. Protokl builds your split around the days you actually train, tracks per-muscle weekly volume as you log, and moves sets between sessions when one workout is carrying too much. You get the distribution right without running a spreadsheet. Try it free for seven days.