Progressive overload, explained with a log

Progressive overload means asking a little more of your body over time, so it keeps adapting: a rep more, a small plate more, a set more. Next time, beat one number from last time. That only works if you know what you did last time, so it starts with a log. To compare sets of different weights and reps, the One-Rep Max Calculator turns each into one number.

The Next-Set Calculator works out the next target from your last session.

What progressive overload means

The American College of Sports Medicine's 2009 position stand defines it in one line: "Progressive overload is the gradual increase of stress placed upon the body during exercise training." In plain words: last month's work has to get a little harder.

The same stand puts the reason beside it: "Systematically increasing the demands placed upon the body is necessary for further improvement". Both lines are in the 2009 stand on progression models.

ACSM replaced that stand in 2026 with a review of 137 systematic reviews covering more than 30,000 participants. Its conclusion begins with four words: healthy adults "should perform progressive RT" (resistance training). The 2026 stand also found that few details of a program changed the main results, and that strength gains were greater with heavier loads, a full range of motion, two to three sets, and at least two sessions a week. After 137 reviews, the instruction keeps the same word: progressive.

Five ways to add load, ranked by how easy they are to write down

The 2009 stand lists five ways to raise the demand: more weight, more reps at the same weight, a different speed per rep, changed rest periods, and more total work, raised gradually by 2.5–5%. All five count. They are not equally easy to record, so the order below is ours, by how easy each is to write down and check in a log.

The stand's own wording on rest is worth keeping straight: shorten rest periods for endurance work, or lengthen them for strength and power. The ranking above is ours, not ACSM's.

The top two rows are the ones to use most weeks. Each is a single number you already write down, and one look at last session tells you whether it went up. Rows four and five ask you to measure something that needs a clock: the seconds between sets, the seconds per rep. A number you do not write down cannot be beaten, because you cannot remember what it was.

Is it only adding weight?

No. Reps count too, and two studies found little between the two methods. A 2022 study of 43 trained lifters compared adding load with adding reps over eight weeks and called the differences "of questionable practical significance". A 2024 study of 39 untrained adults trained one leg each way for ten weeks and found no difference. Weight is one way; reps are another.

Plotkin and colleagues, in PeerJ, put those 43 trained lifters through eight weeks of lower-body training twice a week. One group added load and kept the reps constant; the other added reps and kept the load constant. Muscle growth leaned slightly to the reps group, strength slightly to the load group. Their conclusion: both "appear to be viable strategies" over an eight-week cycle.

Chaves and colleagues, in the International Journal of Sports Medicine, trained 39 previously untrained young adults for ten weeks, one leg progressing by load and the other by reps on the leg extension. Strength rose from about 52 kg to about 68 kg on both legs and muscle size rose on both, with no difference between the two methods.

Two studies, two settings, no clear winner between load and reps. The practical use of that is simple. The smallest step your gym's plates allow is sometimes a bigger jump than the next session can take. When it is, add a rep at the same weight and come back to the plates later.

One month in a log

Here is a month of squats written down: three sets, a range of 6–8 reps, twice a week, 2.5 kg steps. The weight does not move until all three sets reach 8. For six sessions the bar stays at 60 kg while the reps climb, and the log still shows the work going up, in total kilos and in estimated max.

"Total lifted" is the weight times the total reps: 60 kg × 18 reps = 1,080 kg in session 1. "Estimated max" is Epley's formula on the best set, weight × (1 + reps ÷ 30). So 60 kg × 6 gives 72.0 kg, 60 × 7 gives 74.0, 60 × 8 gives 76.0, 62.5 × 6 gives 75.0 and 62.5 × 7 gives 77.1. It is an estimate from one set, not a tested max, which is its own subject: are one-rep max calculators accurate?

From session 1 to session 6 the weight never changed. Total reps went from 18 to 24, total work rose by a third, from 1,080 kg to 1,440 kg, and the estimated max rose 4 kg, from 72.0 kg to 76.0 kg. Six sessions of progress with the same plates on the bar.

At session 7 the bar went up 2.5 kg and the reps dropped back to 6, 6, 6. Total work fell to 1,125 kg, well under session 6's 1,440 kg, and the estimated max dipped 1 kg to 75.0 kg. One session later, 7, 7, 6 at 62.5 kg gives 1,250 kg and an estimated max of 77.1 kg, the highest figure of the month.

The example waits for all three sets to reach 8 before the weight moves. That is one common, strict version. Masheen's coach reads only your best set and the reps you had left.

How fast should the numbers go up?

The one sourced number: the 2009 ACSM stand recommends raising the load by 2–10% once you can do one to two reps over the target number on two consecutive sessions, the lower end for small-muscle exercises and the higher end for large ones. On a 60 kg squat, 2.5 kg is 4.2%, which sits inside that band.

The trigger matters as much as the size. Two sessions running, not one. A single good set says you had a good set.

The same stand notes that among untrained people adaptations "may occur in a short period", and that the size of improvement depends on training status. If you are in your first four weeks in the gym, that is the part to know.

How to keep the log

Four things per set: the exercise, the weight, the reps, and how many reps you had left. The first three say what you did. The fourth says what it cost. Write them while you rest, not afterwards from memory. A log you can read back in ten seconds is worth more than a full one you never open.

Reps left, sometimes called reps in reserve, means how many more you could have done. It is the column that earns its place. Two sets of 8 at 60 kg are not the same set if one had three reps left and the other had none, and only one of them says "add weight next time". That is the number double progression runs on.

Masheen is a workout log for people who lift. It is coming soon to iPhone and Android. It shows what you lifted last time on the set screen, so the number to beat is in front of you. After each set it works out your estimated one-rep max with Epley's formula and draws the trend over time. Its coach suggests the next set by double progression and shows the reason, and you can see the coach on the Masheen home page. The coach and strength trends are part of Masheen Pro; logging is free.

When the numbers stop going up

A stall is the log telling you something: the best numbers on one exercise stop moving. Masheen's coach treats three sessions with no new best estimated max on an exercise as a stall, and then suggests dropping about 10% next time and building back up, or swapping in a close variation for a few weeks. That is one rule of this kind, not the only one. The point is that a written rule reads the log for you.

If the weight will not move, Table 1 still has rows. Hold the weight and go for a rep: that number is as easy to record as the first. A stall on one exercise is also not a stall on the rest, and the log is what tells you which is which. The rule that uses the reps-left column is set out in double progression.