One-Rep Max Calculator

Every rep-max equation assumes the set you are reporting ended at or very close to failure. That assumption is where most bad estimates come from, because most sets in most gyms stop earlier than the lifter thinks. The fix is simple: feed in the reps you could have done, not the reps you did. Here is what that is worth, taking a set of 8 reps at 100 kg:

The pattern is close to linear: each rep left in reserve costs about 2.9% of your estimate, so stopping 4 reps short understates your max by roughly 12%. Nothing about the formulas is wrong here — they were simply given the wrong number.

The standard way to keep track of this is the reps-in-reserve scale validated by Zourdos and colleagues, where an RPE of 10 means no reps left, 9 means one left, 8 means two, and so on. Logging an RPE next to every set costs nothing and makes your estimates comparable from month to month. In that study, experienced squatters and novices rated the same relative loads differently at the top end, so expect your own sense of "one rep left" to sharpen with practice.

Most structured strength programs express working weights as a percentage of your one-rep max. Heavy strength work typically lives at 85–95% for low reps, hypertrophy (muscle-building) work around 67–80% for moderate reps, and lighter technique or endurance work below that.

The percentage table turns your estimate into concrete loads for each of these zones, so you can pick the weight that matches the set and rep scheme your program prescribes. Round to whatever your available plates allow.

That rounding is worth a moment's thought, because it sets the real precision of the whole exercise. With 1.25 kg plates, the smallest change you can make to a barbell is 2.5 kg — about 2.0% of a 125 kg max — and one step down the percentage table is 6.2 kg, which is two or three of those increments. Anything finer than that is arithmetic you cannot load. When a prescribed weight falls between plates, round down: the missing kilogram costs you nothing, while the extra one is paid for across every set of the block.

There is direct evidence for how short the set should be. Reynolds and colleagues put 70 adults through 1, 5, 10, and 20-rep maxima on the bench press and the leg press, then tested how well each could predict the tested single. The 5-rep sets predicted best (R² of 0.993 on the bench press and 0.974 on the leg press), accuracy fell at 10 reps and again at 20, and the authors concluded that no more than 10 reps should be used in a linear prediction equation at all. They also found the load-to-rep curve is not straight, and that the bench press and the leg press needed their own equations — which is why one chart cannot fit every lift.

• Use a lower rep set. Estimates from sets of 2–5 reps are the most accurate. The higher the reps, the more individual endurance skews the result.
• Only count clean reps. Include reps done with good form through a full range of motion — not grinding, half-range, or assisted reps.
• Base it on a near-limit set. The relationship assumes the set was genuinely challenging. A weight you stopped well short of failure with will underestimate your max.
• Estimate each lift separately. Your 1RM differs by exercise, so use a set performed on the specific lift you care about.

An estimate you never test drifts. The cheapest way to check one is to let a working set audit it, which takes no extra training and no maximal attempt:

1. Take your estimate and load the 85% row of the percentage table. For a 124.8 kg estimate that is about 106.0 kg.
2. Do one set to roughly one rep in reserve, counting honestly, and log the reps.
3. Compare with what the table expects at that load — five or six reps. Landing in that window means your estimate is sound.
4. If you are well outside it, put the actual set back through the calculator. Getting 8 reps at 106.0 kg implies a 1RM of 132.3 kg — about 6% above where you started.

Re-run that check every four to six weeks, or whenever the prescribed loads start feeling wrong in either direction. Strength moves faster than most programmes assume early on, and an estimate three months old is usually the reason a block feels either trivial or crushing.

A tested single is the only way to know the number rather than estimate it, and it is the one part of this page that carries real injury risk. It is worth doing only if you are experienced with the lift, and only on a day when you are fresh — a max tested at the end of a hard week measures your fatigue, not your strength. Conventional practice looks like this:

• Set up the safety first. A spotter for the bench press, or pins and safety bars set at the depth where you would need them. Decide before you start that a rep which stalls gets dumped rather than fought.
• Ramp in singles, not sets. Something like 50, 70, 80, 90, then 95% of your estimate, one rep each, with two to four minutes of rest. The warm-up should sharpen your technique, not accumulate fatigue.
• Give yourself two attempts, not five. Take a first attempt near your estimate. If it moves well, add a small jump and stop there. Repeated near-misses cost more than the information they buy.
• Stop at form breakdown. A rep that changes shape is the end of the test, whatever the plates say. Log the load and the rating, and go back to normal training the next session.

For most people, most of the time, a hard set of five and this calculator will give a number good enough to programme from, at none of the cost.