What a 1RM is and how to use it without testing a max

RM stands for “repetition maximum”. Your squat 1RM is the weight you can do 1 full rep with and not a second one. Your 5RM is the weight you can do exactly 5 reps with. It's a number for each exercise: your bench press 1RM tells you nothing about your squat.

The American College of Sports Medicine (ACSM) updated its 2009 position stand in 2026 with a summary of 137 systematic reviews covering more than 30,000 healthy adults^[1]. Its recommendations apply both to people who have never trained and to those with years of experience, and it concludes that many different programmes work^[1]. Load matters mainly for strength, which improved more with 80% of 1RM or more; muscle size didn't depend on load^[1]. To raise the weight as you get stronger, the ACSM considers regular strength testing or perceived exertion scales^[1]. Perceived exertion scales don't require knowing your 1RM.

Lifting close to your max also sends blood pressure soaring. In a 1985 study of 5 experienced bodybuilders, with pressure measured inside the artery, the leg press to failure took the group average to 320/250 mmHg, and in one of them it went above 480/350^[2]. Part of that rise came from holding the breath (the Valsalva manoeuvre), which showed up both in a maximal lift and when nearing failure with a lighter weight^[2]. It's only 5 people, but it helps explain why a max isn't something to test lightly.

To estimate it without testing it there's the 1RM calculator: you enter a weight and the reps you do with it, and it gives you the estimate and a table of loads. Its page explains how it calculates.

The formulas start from one idea: the more reps you can do with a weight, the further that weight is from your max. The relationship isn't a straight line. Reynolds et al. tested 70 people (34 men and 36 women, aged 18 to 69) on the bench press and the leg press: the weight drops non-linearly as reps go up, the most accurate estimate came from sets of 5, and they recommend using no more than 10 reps in linear formulas such as Epley's^[3].

Whether a formula gets it right depends on the exercise and on who does it. In 20 trained people who did the bench press to failure at 70% of their 1RM, Epley's had the lowest standard error, against 2 other formulas and a velocity-based estimate^[4]. In 18 people with knee osteoarthritis, Epley's was among the most accurate on the knee extension, with typical errors of 3 to 4%. On the leg press, however, other formulas did better, with typical errors of 5.9 to 6.3%^[5].

These are small studies, with one or two exercises each. Your estimated 1RM is good for starting to choose loads, and it gets corrected by what happens in your sets.

The same percentage of your 1RM doesn't give the same reps on every exercise. Shimano et al. tested 16 men (8 trained and 8 untrained) at 60, 80 and 90% of their 1RM: they did more reps on the squat than on the bench press or the biceps curl, especially at 60%, and being trained made little difference, except on the bench press at 90%^[6].

The kind of training you do counts too. On the leg press at 80% of 1RM, 8 distance runners did an average of 19.8 reps, and 8 weightlifters, 11.8^[7]. At the same percentage, some did about 8 reps more than others.

That's why the percentage gives you the weight for your first set, and the reps you get tell you whether it was right. To keep progressing in the long term, the ACSM says you need to increase the stimulus, even if only slightly, and raising the load is one way to do it^[1]. If you're just starting out, learn the movements first with a light weight.