The question arrives in a specific form ("how many sets per muscle per week?") and it wants a number, and there is one, and it will not help you as much as you hope. Here it is anyway, because withholding it would be coy:
Ten to twenty hard sets, per muscle group, per week. Beginners do well at the bottom of that. Most trained lifters live in the middle. The upper end is where the evidence gets thin and the arguments get loud.
Now the part that decides whether the number is any use.
Nobody knows their own current number
Ask a lifter how many sets of chest they did last week. You will get an answer, delivered promptly, and it will be wrong, usually low, and the size of the error is frequently a third.
Because volume is not counted where it accumulates. You count exercises. Bench press, four sets; fine, that is chest. Then incline dumbbell press, three sets, also chest. Then a machine fly, three sets, chest again. Then Friday's overhead press, which you filed under shoulders and which your upper chest and triceps also attended. And the dips, which you did not really count because they were at the end and you were leaving.
Everything lands on the muscle, not in the category you filed it under. Add it up honestly and the man who was worried he might be under-training his chest at "eight sets" is doing eighteen and wondering why his shoulder aches.
The mirror error is equally common: someone convinced they are training legs hard, doing four sets of squats a week and nothing else, and describing the resulting fatigue as evidence of high volume when what they have is high intensity and very little else.
This is why the number does not settle the argument. Two people using the same target are running programs a factor of two apart, and both believe they are compliant.
Count what you actually did for six weeks before changing anything. Then the target becomes usable. It is the same instruction as logging your food before touching your calories, and it is unpopular for the same reason: it is a week of work that produces no visible progress, only accuracy.
What counts as a set
Not everything with a number beside it.
Warm-ups do not count. Obviously, but people include them when they are self-reporting and exclude them when they are complaining about fatigue.
Sets taken close to failure count. Somewhere in the last few reps you should be genuinely uncertain about the next one. A set of twelve you could have taken to twenty is a set your muscle largely ignored, and the log has no way to know that unless you write down how hard it was.
This is what an RPE field is for, and it is the highest-value optional number in the whole log. Eighteen sets at RPE 6 and twelve sets at RPE 9 are not comparable weeks, and without that column they look identical, which will send you looking for the cause of your stall in every place except the one it is hiding.
Indirect work counts partially. Triceps get worked by every press, biceps by every row. Nobody has a clean conversion factor and anybody offering you one is selling something, but "some" is very much closer to right than "none," and the people counting it as zero are the people with sore elbows.
More is not free
The dose-response curve flattens. Where it flattens is personal: it depends on the muscle, your training age, your sleep, your food, and how much of your recovery is already being spent on a job and small children and the other things that never appear in a training study.
Past your own flattening point, extra sets buy fatigue at full price and adaptation at a discount. That fatigue does not vanish. It shows up four weeks later as a top set that grinds, a joint that complains before the muscle does, and a sudden inexplicable reluctance to train. Which is the deload conversation, arriving on schedule.
And time. Twenty-five sets a week per muscle is an enormous amount of gym for a return that may be a rounding error. There is a version of this hobby that consumes your entire life for a marginal improvement, and plenty of people are living in it.
How to actually adjust it
Change one variable, slowly, and give it long enough to testify.
Add two to four sets per week to the muscle group you care about. Hold it for six to eight weeks. Then look at whether the working weights moved and whether recovery held. If yes, keep it. If the lifts stalled and everything aches, you overshot and now you know something specific about your own capacity that no article could have told you.
What you cannot do is change volume, exercise selection and calories in the same fortnight and then attribute the result to any of them. That is not a training block, it is a weather report.
Doing it in OneRep
Hard sets per week, counted from what you actually logged rather than what you meant to do, and compared against the week before, so the creep is a number instead of a suspicion. Every exercise carries its primary muscles, which is the tedious half of the tally already done.
Log RPE per set if you want the volume number to mean anything, then read strength trends and volume on the same timeline as your calories and body weight, because a stall in a deficit is a different diagnosis entirely and adding four sets will not fix it.
Ask Coach whether your volume is doing anything. It has the sets, the trend and the intake in front of it, which is a better starting position than the one you have at 6 a.m. trying to remember Thursday.
Try it: Free, no trial window.
Compare: OneRep vs Hevy · OneRep vs Strong