One Rep Max Calculator

How to use this tool

Pick the lift, type the weight you had on the bar and how many reps you got with it, and the number at the top is your estimated one rep max. The Formula box is set to the median of all seven published equations, which is the setting to leave alone unless you have a reason to prefer one. The green panel turns whatever number it gives into a plate list, so you can see what to actually load.

The set has to have been a real one. These equations assume you stopped because you could not do another rep, with the same technique you would use for a single. A set with three reps left in reserve reads about ten percent low, and a set where the last two reps turned into a different movement reads high by more than that. If you are logging with a reps in reserve scale, only a set at RIR zero or one belongs in this box.

The two charts are the reason this page shows every formula rather than one. The left one is your set through all seven. The right one plots them all across every rep count, and the shape of that picture is the practical advice: at three reps they sit almost on top of each other, and at fifteen they are thirty three percentage points apart. If you want a number you can train from, test at five or fewer.

Below that, the percentage table reads the curve backwards. Given your estimated max, it gives the weight for every rep count from one to fifteen, rounded to the plate increment you chose, with what goes on each side of the bar. Every row is a link, so you can send someone a specific working weight rather than describing it.

What the formulas actually are

All seven are regressions: somebody collected sets of data where lifters did a known weight for a known number of reps and also had a tested one rep max, then fitted a curve. They differ in the shape of curve they chose. Epley, published in 1985, is a straight line: multiply the weight by one plus reps divided by thirty. O Conner is the same idea with a gentler slope, two and a half percent per rep. Both are easy to do in your head, which is most of why they survived.

Brzycki, from 1993, and Lander, from 1985, put the reps in the denominator: weight times 36 over 37 minus reps in Brzycki, weight over 1.013 minus 0.0267 times reps in Lander. That shape fits real data better in the middle of the range and then misbehaves badly at the end, because as reps climb the denominator shrinks towards zero and the estimate climbs towards infinity. Brzycki reaches division by zero at exactly 37 reps.

Lombardi, from 1989, uses a power curve, weight times reps to the power 0.1, which flattens out instead of exploding. Mayhew and Wathan both use exponential fits with four constants each, developed on college football players in the early nineties. Mayhew was fitted so tightly to multi rep bench data that it does not pass through the weight you lifted when you feed it a single: it reads about nine percent high there, which is a good reminder that these curves were never meant to be evaluated at one rep.

Why the rep count matters more than the formula

Take a 225 lb bench set. At three reps the seven formulas return between 238 and 256, a span of eighteen pounds, and most of that gap is Mayhew on its own: drop it and the other six sit inside thirteen. At ten reps they return between 281 and 303, a span of twenty two. At fifteen they return between 295 and 368, a span of seventy three pounds, which is not an estimate at all, it is two different answers.

There is a physical reason as well as a mathematical one. A set of three is limited by how much force you can produce. A set of fifteen is limited by how long you can keep producing it, which brings in breathing, blood flow, grip and simple willingness to stay under the bar. Those things vary enormously between people and none of the equations know anything about them. Two lifters with the same true max can differ by five reps at seventy percent.

That variation runs by lift too. Most people get more reps at a given percentage on the squat and leg press than on the bench, so multi rep estimates read low on the lower body. Deadlifts usually go the other way and give fewer reps than the formulas expect, because grip and lower back fatigue end the set before the legs are finished. If you use the same equation across all your lifts, expect it to be biased differently on each of them, and calibrate it against your own tested numbers once.

Percentage of one rep max, by formula and rep count

Exact arithmetic for every formula. Read down a column to see how one equation behaves; read across a row to see how far apart they are at that rep count. The median column is what this page reports by default.

Estimated one rep max as a percentage of the weight lifted. Multiply your weight by the figure and divide by 100. Each rep count in the first column opens the calculator on a 225 lb bench set at that rep count.

Turning an estimate into a training plan

The usual reason to want a one rep max is to set working weights, and the percentage table is the tool for that. Heavy strength work sits around 85 to 92 percent, which is triples and doubles. The classic five rep work sits near 85. Hypertrophy work runs from about 65 to 80 percent, which is eight to fifteen reps. Those bands are conventions rather than laws, and the percentage that corresponds to a given rep count is exactly the curve this page draws.

Two practical notes. First, round to the plates you own rather than to the number the calculator gives: an 8 lb difference on a bench press is noise, and hunting for fractional plates is not worth the time. The plate panel here rounds to whatever increment you choose and shows the loading for each side. Second, most programmes are better run off a training max of around ninety percent of your estimate, because it gives you somewhere to go when a session is bad and stops one missed rep from derailing a week.

If you are using a rate of perceived exertion scale instead, the translation is simple: RIR zero is the percentage this table gives for that rep count, and every rep you leave in reserve is worth roughly two and a half to three percent. A set of five at RIR two is about the same load as a set of seven at RIR zero, which is around 78 percent rather than 85.

A worked check you can repeat on paper

Start with 225 lb for 5 reps. Epley is 225 times one plus five over thirty, which is 225 times 1.16667, which is 262.5. Brzycki is 225 times 36 divided by 32, which is 253.1. Lander is 225 divided by 1.013 minus 0.133562, which is 225 over 0.879438, which is 255.8. Lombardi is 225 times 5 to the power 0.1, which is 225 times 1.17462, which is 264.3.

Mayhew is 225 divided by 0.522 plus 0.419 times e to the power minus 0.275, which is 225 over 0.84034, which is 267.8. O Conner is 225 times 1.125, which is 253.1. Wathan is 225 divided by 0.488 plus 0.538 times e to the power minus 0.375, which is 225 over 0.85786, which is 262.3. Sorted, those are 253.1, 253.1, 255.8, 262.3, 262.5, 264.3, 267.8, so the median is 262.3.

Now read it back. Eighty six percent of 262.3 is 225.6, which is the weight we started from, so the arithmetic closes. Seventy five percent is 196.7, which rounds to 195 lb, and on a 45 lb bar that is 75 lb a side: a 45, a 25 and a 5. That is the whole tool in three lines.

Mistakes that catch people out

Estimating from a set you stopped early. The formulas assume the set went to or near failure. Leaving three reps in the tank understates your max by around ten percent, which is enough to make a whole training block too light.

Using a high rep set because it felt impressive. Twenty reps tells you about your conditioning. Use three to five if the number is going to set your training loads.

Comparing estimates across formulas as if one is right. They are different curve fits to different data sets. A 12 lb disagreement between Epley and Brzycki is not an error to resolve, it is the width of the uncertainty.

Treating an estimate as a competition entry number. Estimates are for choosing loads. If you need the real figure, work up to a tested single with full rests on a day you are fresh.

Forgetting the bar. The plate list on this page subtracts the bar before splitting the remainder, which is what the bar weight box is for. An Olympic bar is 20 kg or 45 lb, but many gyms have 15 kg womens bars and 35 lb training bars, and technique bars can be lighter still.

Recalculating after every session. A max does not move week to week in any way an estimate can detect. Re estimate every four to six weeks, or when a working set moves noticeably faster than it used to.

Popular one rep max lookups

The sets people look up most. Each opens the calculator on that lift, weight and rep count with its own address, so you can bookmark or share the exact answer.

Bench press

Squat and deadlift

Overhead press and rows

In kilograms

By formula

How this was checked

Every equation was entered from its published form and checked in Node before the page was built. The two identities documented in the literature both hold: Epley and Brzycki return exactly the same figure at ten reps, and Epley reads higher than Brzycki below ten and lower above it. The one rep behaviour of each was checked and recorded rather than hidden: Brzycki and Lombardi return exactly the weight lifted, Wathan reads 1.3 percent high, Lander 1.4, O Conner 2.5, Epley 3.3 and Mayhew 8.9, which is why this page returns the weight itself at one rep instead of a formula output.

The breakdown points were checked: Brzycki divides by zero at 37 reps and Lander at 37.9, and both are guarded so the tool refuses rather than printing a nonsense figure. The percentage table was checked to be the exact inverse of the estimate, so feeding a max back through it returns the weight the estimate came from. Plate loading was checked to sum correctly for both plate sets, pounds and kilograms, including the case where the target cannot be made with standard plates. The live page was then opened at every address linked above and each was checked for a distinct title, the right headline, charts that rendered, and clean CSV, image and print output containing only that set, at desktop, tablet and phone widths.

Common questions

How do you calculate a one rep max from a set?

Every method is the same shape: take the weight you lifted, multiply it by a factor that grows with the number of reps you got. Epley uses weight times one plus reps over thirty, so 225 lb for 5 reps is 225 times 1.1667, which is 262. Brzycki uses weight times 36 over 37 minus reps, so the same set is 225 times 36 over 32, which is 253. Seven such equations are in common use and this page runs all of them, reporting the median rather than picking a favourite. The important part is not which one you use but how many reps the set was: at three the answers cluster, and past twelve they scatter.

Which one rep max formula is the most accurate?

None of them is reliably the best, which is why this page shows all seven. Epley and Brzycki are the two most used and they return identical answers at exactly ten reps, with Epley reading higher below that and lower above it. Brzycki and Lander climb steeply at high reps because both divide by a shrinking number, and Brzycki breaks outright at 37 reps where its denominator reaches zero. Lombardi is the most conservative at high reps by a wide margin. The practical answer is that the choice barely matters under six reps and matters enormously over twelve, so change the set rather than the formula.

Is a one rep max estimate accurate enough to train from?

For choosing working loads, yes, provided the set was a hard one. Estimates from sets of three to six reps typically land within a few percent of a tested max for trained lifters, which is well inside the resolution of the plates you own. For anything where the exact number matters, like a competition attempt or a record, test it: work up in singles with long rests and find the real figure. The estimate is a tool for programming, not a result.

Why does the calculator give my exact weight back at one rep?

Because that is the honest answer: if you lifted it once, that is your one rep max and no estimate is needed. Most of the published equations do not do this. Feed Mayhew a single and it reads about 9 percent high; Epley reads 3.3 percent high; only Brzycki and Lombardi pass through exactly. That is not a bug in those equations so much as a sign of what they are: curve fits to multi rep data, extrapolated to a point they were never really aimed at. This page short circuits the formula at one rep rather than reporting a number you know is wrong.

How many reps should I use for the most reliable estimate?

Three to five. Below that you are close to testing anyway, and above six the equations start to separate and the set stops being limited purely by strength. At fifteen reps the seven formulas span thirty three percentage points for the same set, which means the estimate is a range rather than a number. Breathing, grip and general conditioning start deciding how many reps you get, and none of the formulas know anything about those.

Does the estimate work the same for squats and deadlifts as for bench?

Not quite. These equations were developed largely on bench press data, and rep performance differs by lift. Most lifters get more reps at a given percentage on the squat and leg press than on the bench, so multi rep estimates tend to read low there. Deadlifts often go the other way because grip and lower back fatigue cut the set short before the legs are done, which makes a rep max estimate read low as well but for the opposite reason. Use the same formula across lifts if you like, but calibrate it against your own tested maxes before trusting it to the pound.

Related Calculators and Reference

• Macro Calculator: turns a calorie target into grams of protein, carbohydrate and fat, with protein set from body weight and the research.
• TDEE Calculator: the calories that support the training, and the intake for a cut or a gain around it.
• Weight Converter: kilograms to pounds, for reading a kilo loaded bar in pounds and back.
• Percentage Calculator: for working percentages of a max by hand, and for the week to week changes in a programme.
• Steps to Miles Converter: for the conditioning side of the week, turning a step count into distance and time.
• Pace Calculator: pace, time and distance for the running that sits alongside the lifting.
• Protein Intake Calculator: grams of protein a day for your weight and goal.