Tracking

How to estimate your one rep max, and why you rarely need to

Most strength calculators ask the same thing: what weight did you lift, and for how many reps? Type in 100 kg for 8 and one tells you your max is 127 kg. Another says 124. A third says 120. Same set, three answers, and no explanation of why.

The explanation is short. Every estimate is a formula, every formula is a curve fitted to how people's reps fall as weight goes up, and those curves agree in some places and disagree badly in others. Once you know where, you can use them properly, and you'll also see why you need them less often than calculators suggest.

The short version

• One rep max formulas turn a multi-rep set into a guess at the heaviest single you could lift.
• They agree reasonably well at low reps and split apart quickly above 10.
• Research suggests using sets of no more than about 10 reps to estimate.
• If you didn't go to failure, add your reps in reserve to the rep count first.
• To track progress, compare sets at the same weight instead. No estimate needed.

How the formulas work

Two formulas cover most of what you'll see online.

Epley multiplies the weight by one plus the reps divided by 30. A set of 100 kg for 5 becomes 100 × (1 + 5/30), about 117 kg.

Brzycki multiplies the weight by 36 and divides by 37 minus the reps. The same set becomes 100 × 36 / 32, about 113 kg.

There are several others, with names like Lander, Lombardi and O'Conner. Each gives a slightly different answer from the same set, which is the whole reason three calculators can hand you three numbers.

All of them are just curves describing a pattern everyone has felt: each extra rep you can do at a weight means you could lift a bit more for one. They disagree about how steep that curve is, and the disagreement depends on where you are on it.

Where they agree, and where they don't

Run 100 kg through both at different rep counts and the pattern is obvious:

• 3 reps: Epley 110, Brzycki 106.
• 5 reps: Epley 117, Brzycki 113.
• 10 reps: both 133.
• 15 reps: Epley 150, Brzycki 164.
• 20 reps: Epley 167, Brzycki 212.

At low reps the gap is a few kilos. At 10 they happen to cross. Past 12 they come apart fast, and at 20 one formula claims you could lift 45 kg more than the other. Neither is "right". They're both extrapolating far from the data that shaped them.

The research points the same way. Reynolds, Gordon and Robergs tested 70 adults on bench and leg press at 1, 5, 10 and 20 rep maxes. The relationship between load and reps wasn't a straight line, the 5-rep max gave the best prediction, and they recommended using no more than 10 reps in linear equations. A study of 103 women by Mayhew and colleagues found the same thing for bench press: equations were more accurate when fewer than 10 reps were used, before and after 12 weeks of training.

So the practical rule is simple: estimate from a heavy set, not a long one.

The set has to be hard, or you have to say how hard it was

These formulas are meant for sets taken to failure, and studies testing them, like the ones below, had people go to failure. If you stopped at 8 with 3 more in you, plugging in 8 tells the formula you're weaker than you are.

The fix is to add the reps you had left. That gives you the set's effective reps, which is reps plus reps in reserve. Eight reps with three in reserve is an 11-rep effort as far as the formula is concerned. We explain the idea fully in what an effective rep is.

That fix has its own error bar. Lifters tend to underestimate how many reps they have left by about one on average, according to a meta-analysis by Halperin and colleagues, and accuracy is worse on longer sets. So a sloppy reps-in-reserve guess on a long set stacks two errors on top of each other. Another reason to estimate from heavy, short sets.

Why the same formula fits some people better than others

One more thing the calculators don't tell you: the rep-to-load curve isn't the same for every person. A small study by Richens and Cleather had eight weightlifters and eight endurance runners do as many leg press reps as they could at 70 percent of their max. The runners averaged about 40 reps. The weightlifters averaged about 18. Same percentage, more than double the reps.

Run either group's numbers backwards through one formula and you'd get very different estimates of a max that was actually measured. A formula is an average curve, and you are not an average. Your training history bends the curve, and the formula has no way of knowing it.

The lift matters too. In the Reynolds study, the error of their own 5-rep prediction was about 3 kg on the bench press and about 16 kg on the leg press. So an estimate can be decent on one lift and noticeably off on another. If you've ever compared a calculator's guess with a real max and seen it miss by a lot, this is one reason. It's also part of why strength standards calculators give different answers: they're often estimating your max before they even look up a table.

Why you rarely need a true max

A one rep max is a useful number in two situations: you compete in a strength sport, or your program sets loads as percentages of it. Outside those, it's mostly a curiosity, and chasing it has costs. A max attempt is a heavy, fatiguing single that tells you about one day.

For the question most people actually care about, "am I getting stronger", there's a better tool that needs no formula at all. Compare like with like. Same exercise, same weight: did you do more reps, or the same reps with more left in the tank? If yes, you got stronger at that lift, and you didn't have to estimate anything to know it.

That comparison is direct, it uses sets you're already doing, and it doesn't import a formula's error into your record. When the weight changes, the comparison changes too, which we cover in what counts as progressive overload.

There's a quieter benefit to skipping the estimate. Formulas tempt you to compare sets that aren't comparable: a set of 12 last month against a set of 5 today, turned into two maxes and lined up as if they were the same measurement. They aren't. Each one carries its own formula error, and the error is larger for the longer set. The difference between them can be mostly noise.

Comparing sets at the same weight avoids that entirely. It doesn't need anything to be converted, so nothing is lost in the conversion.

The honest concession: an estimated max is still handy for one thing, which is a rough sense of where you stand against other lifters or against your own past self on a different rep scheme. For that, a heavy set of 3 to 5 taken close to failure, run through either formula, is good enough. Just don't treat the decimal places as real.

A formula guesses your strength. Your log measures it.

Common questions

How do you calculate your one rep max?

Take a hard set, ideally of 10 reps or fewer, and put the weight and reps into a formula such as Epley, which multiplies the weight by one plus the reps divided by 30. The result is an estimate, and it is most reliable from heavier sets with fewer reps.

Which one rep max formula is most accurate?

No single formula wins for everyone and every lift. What the research does agree on is that all of them get less reliable as reps rise, with studies recommending sets of no more than about 10 reps for the estimate.

Can I estimate my max from a set I did not take to failure?

Yes, if you add the reps you had left. A set of 6 with 2 in reserve is roughly an 8-rep max, so use 8 in the formula. If you guess the reserve wrong, the estimate will be off by roughly what that many extra reps are worth.

Do I need to test my one rep max?

Only if you compete or follow a program that sets loads as a percentage of it. For tracking progress, comparing your sets at the same weight over time is more direct and avoids the error the formulas add.