Pick a training split and it feels like the programming question is settled. Here is what an upper/lower split is, and then the more useful thing: exactly where its job ends.
What is an upper/lower split?
An upper/lower split divides your training week into two session types: one that trains everything above the waist, one that trains everything below it. You alternate them. On four days a week, that is two upper sessions and two lower sessions, so both halves of the body get trained twice.
A typical week:
The days are not fixed. Monday, Tuesday, Thursday, Friday is common because it puts a rest day in the middle and leaves the weekend free, but Tuesday, Thursday, Saturday, Sunday delivers the same thing. What matters is the alternation and the count, not the calendar.
That is the whole structure, and it takes one diagram to hold. Now the part that takes longer: Monday still arrives with no weight on the bar, no set count, and no answer to whether any of it should differ from last Monday. The split did not decide a single one of those, and that is not a flaw in it. It is what a split is.
Why a second session per muscle group
Two numbers frame it, and both are guidance for getting more out of your training rather than thresholds you fall off. Muscle grows on less than either: ACSM’s own headline finding is that the biggest jump in the whole field is from no training to any training at all. Its guidance puts the productive range for growth at around ten hard sets per muscle per week and up, flattening out past roughly eighteen to twenty. The IUSCA hypertrophy position stand recommends about ten sets per muscle as the most one session is likely to use well.
Around ten a week, those two do not pull against each other at all. Ten weekly sets sit inside what one session can use, and IUSCA says so directly: a muscle can grow trained once a week if the weekly volume is there. A second session starts earning its place as your weekly volume climbs, because fourteen or sixteen sets in one sitting is further past that per-session recommendation than most people want to test. What the evidence says about sets past that point is not that they hurt you. It is that they stop showing a detectable return, which means you spend the time and carry the fatigue and get nothing measurable back for either. That is the argument for splitting, and it is a packing argument, not a frequency law.
The frequency call itself belongs to ACSM alone: train each major muscle group at least twice a week. IUSCA does not make that call, and it is worth knowing the two stands differ here rather than hearing “twice a week” as settled consensus.
One more caveat, because it is overstated in both directions. For muscle size, frequency itself is not the active ingredient: once your weekly set total is fixed, spreading those sets over more days adds little on its own. Strength is where that stops being true. A 67-study meta-regression, circulated as a preprint before publication, found a genuine if diminishing effect of frequency on strength, so a lift you want to get stronger at earns repetition in its own right. The “frequency does not matter” line you will hear repeated is a hypertrophy finding, and it does not travel to strength. The four numbers underneath all of this get their own treatment in how many sets, how close to failure, how often.
Which is the first sign of something this post keeps running into: almost every number above is a hypertrophy number.
If your goal is strength, the arithmetic is different
The ten-in, ten-out reasoning is hypertrophy reasoning end to end. A strength lifter reading it and concluding that ten heavy squat sets in a day is the target has been misled by a number that was never about them.
Two differences matter. First, the set response reaches a functional plateau far earlier for strength than it does for growth. Strength needs fewer hard sets, and buys its progress with load and practice instead. ACSM’s own dose-response puts strength’s flattening point at around two to three sets per exercise in a session.
Second, the ten-set per-session figure is a hypertrophy recommendation. It sits in a hypertrophy position stand and was not written with near-maximal loads in mind, so a strength lifter should not read it as a target. The direction of that correction is worth being clear about: it runs tighter, not looser.
So a strength lifter still splits their week, but not for the volume packing. It is the frequency effect above, plus a coaching argument rather than a cited one: a heavy lift is a skill, and skills reward repetition. A permission comes with that, and it points the same tighter direction: strength gains are largely indifferent to how close to failure you train, so a heavy compound can sit at one to three reps in reserve or further back without costing you the adaptation. The currency here is hard sets near your top end, not sets in total.
Training age moves these too. IUSCA is explicit that how close to failure a lifter needs to train rises with experience, and ACSM’s novice finding is that early gains come from doing resistance training at all, without progressive overload being required yet. Neither stand converts into a weekly set count for a particular lifter, which is rather the point. They describe how a population responds on average, and where any one lifter sits inside that is something only their own training answers.
What the split decides, and what it leaves open
A split decides three things, and they are all scheduling.
Which muscles train on which day. The obvious one, and the only one most people count.
How much rest each area gets between sessions. Three-day full body on Monday, Wednesday, Friday comes back to every muscle at forty-eight, forty-eight, then seventy-two hours across the weekend. Upper/lower on Monday, Tuesday, Thursday, Friday stretches each half to seventy-two, then ninety-six. Same muscle, same weekly volume, a different recovery window between exposures.
Which lift you meet fresh. Every session has one slot at the front where you are least tired, and the split determines what can occupy it. On a three-day full body, only one of squat, bench and deadlift opens each session. On upper/lower you get two upper openers and two lower ones a week. This one matters more than it sounds: exercise order is one of the few structural variables ACSM identifies as a genuine driver for strength, while having no effect for size.
Now the things it does not decide, which are the ones that determine whether the training works.
How heavy. The split does not know your numbers. Load is where the two goals diverge hardest: strength work lives in a heavy zone, size responds across a much broader range of weights as long as the effort is real. A schedule cannot pick for you.
How many sets. Guidance is not a prescription, and where you land inside it depends on your goal, your training age, how you respond, and what you are currently recovering from. The split has a view of none of that. If you have no number of your own yet, start near ten weekly sets per muscle and let what you recover from move it.
How close to failure. ACSM’s effort target for growth is about two to three reps in reserve, meaning you stop with two or three good reps left. Near failure, not at it, and as above, a strength lifter has more latitude here than a growth lifter does. Nothing in an upper/lower schedule enforces that, or notices when you ignored it.
What changes next week. This is the big one. A split is identical in week six and week one. It has no memory, no progression rule, no mechanism for backing off when you are beaten up and no mechanism for pushing when you are not. Week four looks exactly like week one, because a schedule is all it ever was.
That last gap is the one people feel. Running a template for a couple of months and noticing the weights have not moved is not a sign you picked the wrong split. It is a sign that picking a split was never going to do that job.
Upper/lower, push/pull/legs, or full body?
Mostly, this follows from how many days a week you train, and much less from preference than the debate suggests.
At one to three days, full body is the usual answer, and the reason is the per-session recommendation rather than a frequency rule. Divide three sessions two ways and something gets trained once, which means its entire week has to fit inside whatever a single session can use well. Covering the whole body each time spreads the load instead. On the strength goal the same answer arrives by a different road: with only three sessions, full body is what lets every main lift get practised in all of them. At four days, upper/lower packs evenly. Five is the genuinely open one: no division lands evenly on an odd number of sessions, so you are choosing which unevenness you prefer. Alternating upper and lower works, and runs three of one and two of the other, flipping the following week. A mixed arrangement that adds a full-body day is another answer. At six, push/pull/legs runs twice through cleanly.
None of these is superior training. They are different ways of distributing the same weekly volume, and the evidence does not crown one. It is worth noting that the ACSM overview organises its own coverage checklist as upper and lower crossed with push and pull, so the familiar split names are at least cutting along the lines the literature uses to make sure nothing is missed. That is a statement about covering the body, not an endorsement of any particular split.
Which means the popular version of this decision, agonising over which split is best, is agonising over something far less consequential than what you do inside it.
What a program adds
A program starts where the split stops. It carries three things a schedule cannot.
A direction across weeks. Not only what today is, but what today is for, and how it differs from the same session three weeks from now. The clearest example is a deload week: a planned week where the sets come down on purpose, often the bar with them, because the three weeks before it were hard. A split cannot schedule one, because a split does not know a hard three weeks happened. It has no week three.
A next decision that reads the last one. What you lifted, whether you finished the sets, how hard they were. A program that does not read your training is a calendar with opinions. This is what progression means, as distinct from adding weight because it is Monday.
A reason for the number. If the weight went up, something justified it. If it came down, something justified that too, and you should be able to see what. The worked chain from one hard set to a four-week wave is the long version of that idea.
A split can be written on the back of a receipt. A program is the ongoing work of deciding, every session, what the evidence of your training says to do next. That work is real, it is weekly, and it does not stop. Handing yourself a split and calling the programming done is the most common way lifters end up doing the same thing in month four that they did in month one.
How Sterka handles it
Sterka does not ask you to pick a split, because the answer follows from a question it already asked: how many days a week do you train.
As the engine is tuned today, one to three days is programmed as full body, four as upper/lower, five as upper, lower and full body on a rotation that shifts week to week rather than repeating a fixed Monday pattern, and six or more as push/pull/legs. These are programming defaults, not the only arrangements that work at each frequency. Change your weekly frequency and the structure follows it, without a separate decision to get wrong.
What the engine spends its attention on is everything the split left open. Each session’s weights, its set counts, how close to failure each set should be taken, and what all of that becomes next week given what you actually logged. The schedule is the easy half. The engine is built for the other one, and what it tells you about a number is the arithmetic that moved it rather than a summary written afterwards.
References
- American College of Sports Medicine position stand. Resistance training prescription for muscle function, hypertrophy, and physical performance in healthy adults: An overview of reviews. Medicine & Science in Sports & Exercise, 2026.
- Schoenfeld BJ, Fisher JP, Grgic J, Haun CT, Helms ET, Phillips SM, Steele J, Vigotsky AD. Resistance Training Recommendations to Maximize Muscle Hypertrophy in an Athletic Population: Position Stand of the IUSCA. International Journal of Strength and Conditioning, 2021;1(1).
- Pelland JC, Remmert JF, Robinson ZP, Hinson S, Zourdos MC. The Resistance Training Dose-Response: Meta-Regressions Exploring the Effects of Weekly Volume and Frequency on Muscle Hypertrophy and Strength Gain. Circulated as a SportRxiv preprint in 2024 and published in Sports Medicine in 2025.