Full Body vs Split: What Actually Builds Muscle?

The once-a-week body-part split is an inheritance from competitive bodybuilding culture, not from a controlled comparison. For two decades the counter-argument has been that training each muscle twice a week beats once a week — and that claim does have a well-known source. A 2016 systematic review of ten studies by Schoenfeld, Ogborn and Krieger found a significant effect of training frequency on hypertrophy (p = 0.002), with higher frequencies producing a larger effect size than lower ones (0.49 versus 0.30).^[2] That paper is where "train each muscle twice a week" entered the mainstream.

The problem is what happened next. When the same research group repeated the analysis in 2019 with 25 studies rather than ten, the picture changed. Pooling studies in which total weekly volume was matched between the high- and low-frequency groups, they found no significant difference in hypertrophy — and they stated it plainly: there is "strong evidence that resistance training frequency does not significantly or meaningfully impact muscle hypertrophy when volume is equated."^[3] Where an advantage for higher frequency did appear was in the meta-regression of studies that did NOT equate volume, and even there the authors called the magnitude between one and three-plus days per week "modest".^[3] That is the whole story in miniature. The reason a once-weekly body-part split tends to underperform is not that the muscle needs the stimulus twice. It is that training a muscle once a week usually caps how many hard sets you can give it before the session falls apart. Add the frequency check from the protein-timing piece: an unequated comparison answers a different question than the one in its title.

The three real levers: weekly hard sets per muscle, proximity to failure, and progressive overload. Volume: Schoenfeld's 2017 meta-regression of 34 treatment groups across 15 studies found a graded dose-response — each additional weekly set was associated with an effect-size increase of 0.023, corresponding to about 0.37% more muscle gain per set (p = 0.002), and the difference between the higher- and lower-volume arms within studies came to 3.9%.^[4]

Effort: Refalo and colleagues pooled 15 studies on how close to failure sets are taken and found only a trivial advantage for training to set failure over stopping short — effect size 0.19 (95% CI 0.00 to 0.37) — with no advantage at all for momentary muscular failure specifically, 0.12 (−0.13 to 0.37), and none for higher velocity-loss thresholds, 0.08 (−0.16 to 0.32).^[10] A later continuous meta-regression by Robinson and colleagues, modelling repetitions in reserve rather than a yes/no failure split, did find that hypertrophy improved as sets were terminated closer to failure, while strength gains were similar across a wide range of repetitions in reserve — though the authors describe their own analysis as exploratory, note the model fit was modest, and caution that the proximity-to-failure values had to be estimated from study descriptions rather than measured.^[11]

The honest summary: sets have to be hard, you do not have to grind every one to failure, and a set stopped five reps early is probably not doing much. Progressive overload is the third lever and the one with no shortcut — the load, reps or sets have to go up over time.

The most complete analysis available is Pelland and colleagues' 2025 series of Bayesian meta-regressions in Sports Medicine: 67 studies, 2,058 participants (79.1% male, mean age 25.2 years), with every contributing set classified as direct or indirect to the muscle being measured and all models adjusted for intervention duration and training status.^[1] For weekly volume, the posterior probability that more produced more was 100% for both hypertrophy and strength, with diminishing returns in both and markedly steeper diminishing returns for strength. For frequency, the two outcomes separated: the posterior probability for strength was also 100% — more frequent training produced more strength, with diminishing returns — but for hypertrophy it fell short of 100%, which the authors describe as "compatible with negligible effects."

Then the controlled trials. Hamarsland and colleagues randomised 21 trained men and women to the same weekly volume spread over either two or four whole-body sessions for nine weeks and found no difference in any measure of muscle growth or strength; both groups made large gains, and the gains tracked exercise familiarity rather than frequency.^[9] Note the sample is small and unbalanced (13 versus 8), which limits how much weight a null result there can carry.

The practical reading is not "frequency is irrelevant." It is that frequency is a scheduling tool. Spreading twenty sets for a muscle across two or three sessions usually means better quality sets than crushing all twenty in one, and it is far easier to recover from — but the growth comes from the twenty sets, not from the number of days they were spread across.

Baz-Valle and colleagues restricted their review to randomised trials of at least six weeks in trained 18-to-35-year-olds with direct imaging of muscle size, which left seven studies — and found no difference between moderate (12–20 weekly sets) and high (>20) volume for the quadriceps (p = 0.19) or biceps brachii (p = 0.59), though high volume was better for the triceps brachii (p = 0.01). Their recommendation: 12 to 20 weekly sets per muscle group as a standard target for young trained men.^[6] Seven studies is a thin base and the authors say so.

The upper end has been probed directly: Aube and colleagues randomised 35 very strong trained individuals — mean squat of more than twice body mass — to 12, 18 or 24 weekly lower-body sets for eight weeks and found no between-group difference in muscle thickness or regional fat-free mass at all. Strength, if anything, went the other way: a group-by-time trend (p = 0.052) favoured the middle dose, with the 18-set group adding 25.5 kg to their squat against 18.3 kg for 12 sets and 9.5 kg for 24.^[7]

Then the floor. Androulakis-Korakakis and colleagues asked the opposite question — what is the least that still works — and pooled five studies in which trained men performed a single set of 6 to 12 repetitions at about 70–85% of one-rep max, two to three times a week, taken to or near failure, for 8 to 12 weeks. Overall one-rep max rose 12.09 kg (95% CI 8.16 to 16.03), squat 17.48 kg (8.51 to 26.46) and bench press 8.25 kg (0.68 to 15.83).^[12] The authors are careful to call these gains "suboptimal, yet significant," and the confidence interval on the bench press nearly touches zero. The two results together are the useful frame: the floor for getting stronger is remarkably low, the ceiling for getting bigger is higher and individual, and somewhere in between is a weekly set count you can recover from.

Here is the uncomfortable part: there is no randomised trial directly comparing push/pull/legs against upper/lower against full-body at equated weekly volume. What exists compares frequencies, and the closest thing to a split comparison is Schoenfeld and colleagues' 2015 trial, in which 20 well-trained men were assigned either to a split routine training each muscle group once a week or to a total-body routine training everything three times a week, with weekly volume equated. The total-body group gained significantly more forearm flexor muscle thickness; there was no difference in bench press or squat one-rep max.^[8]

But notice what that trial actually varied: not "split versus full-body" as programming philosophies, but once-weekly versus three-times-weekly frequency. It is a frequency study wearing a split's clothes — which is exactly the confound the 2019 volume-equated meta-analysis went on to resolve.^[3]

So what are the splits, mechanically? Full-body trains every major muscle group each session. Upper/lower alternates between the two halves. Push/pull/legs divides by movement pattern — pressing, pulling, and legs — and is usually run twice through in a six-day week, or once through in three days. A body-part split gives each muscle its own day. The first three all let you hit most muscles twice a week at three to six sessions. The last one, run over six days, usually does not — and that, rather than anything about the exercises, is why it tends to come out behind. When a meta-analysis reports "no significant difference between splits," print the interval beside it: these are small, heterogeneous literatures, and an absence of a detected difference in a handful of eight-week trials in young men is not proof of equivalence.

These are illustrations, not prescriptions, and the set counts are arithmetic rather than findings. The decision rule is simpler than the table: pick the structure that lets you reach your weekly set target for each muscle, train most muscles at least a couple of times a week because it makes those sets better rather than because the frequency itself is magic, and above all pick the one you will still be doing in six months.

Canada's 24-Hour Movement Guidelines for adults aged 18 to 64 ask for muscle-strengthening activities using major muscle groups at least twice a week, alongside at least 150 minutes of moderate-to-vigorous aerobic activity.^[13] Two sessions a week is the public-health floor, not the hypertrophy ceiling — but it is worth saying that a two-day full-body habit you keep beats a six-day plan you abandon in March. Recoverable volume varies between people and falls with poor sleep, under-eating and age; if your performance is sliding week over week, the answer is usually fewer sets, not more. Protein intake is the other input that actually moves the needle, and it is covered separately, alongside plant versus animal protein and Peak Nutrition.

This is the distinction the gym argument almost always loses. In the 2025 meta-regressions, frequency had a clearly identifiable effect on strength and a negligible one on hypertrophy.^[1] In the eight-week volume trial, one-rep max rose identically on 6–9 weekly sets and on 30–45, while muscle thickness tracked the dose.^[5] In the proximity-to-failure analyses, hypertrophy improved as sets got closer to failure while strength was flat across a wide range.^[11] Strength is substantially a skill: practising a lift more often, with high-quality sets, makes you better at it, and you do not need enormous volume to get there. Size is a tissue response: it wants accumulated hard work.

If your goal is a bigger squat, more frequent squatting at moderate volume serves you well. If your goal is bigger legs, the weekly set count is the thing to protect. Most people want some of both, which is one more reason the three-to-five-day structures — full-body, upper/lower, push/pull/legs — keep coming out fine. Anyone with a current injury, a joint problem, cardiovascular disease or any medical limitation should have a programme built for them rather than copied from an article.