Skip to content
One-RepMaxCalculator

One Rep Max Calculator With the Formula and Source Behind Every 1RM

A one-rep max (1RM) is the maximum weight a lifter can lift for one repetition in weight training, and this calculator estimates it from any set of 1 to 10 reps with seven published equations.

Estimate a one-rep max from one set
Unit

The default equation is Brzycki (1993): one-rep max = weight ÷ (1.0278 − 0.0278 × reps).

Written by Dr. GideonReviewed by Rohith Singh, Cult Certified Personal TrainerConstants verified

For 100 kg lifted 5 times, Brzycki's equation returns 112.51 kg and the seven equations span 112.5 to 119.01 kg. Every result on this page names its equation, its primary source and how far the seven published equations disagree for the set entered; that disagreement is not a measured error. The calculator then turns the estimate into loads rounded to real plates, rep-max targets, an optional RPE reading and a training-max line. One-RepMaxCalculator is free and needs no account: it runs in the browser and stores nothing. It is funded by Google AdSense and by affiliate links to lifting equipment on the pages whose results call for it; this page carries none, and it sells no programme, coaching, supplement or app. Loads round to IPF competition discs of 25, 20, 15, 10, 5, 2.5 and 1.25 kg on a 20 kg bar, or to the common US gym set of 45, 35, 25, 10, 5 and 2.5 lb on a 45 lb bar, a gym convention rather than a federation rule.

An estimated one-rep max is an equation's prediction of the most weight a lifter could lift once, made from the weight and reps of a set stopped short of it, and printed here with the equation that produced it.

One-rep max estimates from seven equations

A one-rep max is estimated by entering the weight and reps of one set taken close to failure into a prediction equation. This calculator defaults to Brzycki's 1993 equation, 1RM = weight ÷ (1.0278 − 0.0278 × reps), and prints six other published equations for the same set beside it.

Brzycki published the equation in the Journal of Physical Education, Recreation and Dance in 1993 and limited it to sets of fewer than 10 reps taken to fatigue. The other six forms were read in Jiménez and De Paz 2008 and Reynolds, Gordon and Robergs 2006, two peer-reviewed papers that reproduce them; their authors' originals were not available to check.

Table 1. The seven equations, the form each one uses and where that form was read. Brzycki is the proposed default; its author limits it to fewer than 10 reps. The spread beside every result shows how much these equations disagree for one set, not a measured error.
EquationFormWhere the form was readSource grade
Brzycki (1993)weight ÷ (1.0278 − 0.0278 × reps)Brzycki 1993 (original paper)original paper
Epley (1985)weight × (1 + reps ÷ 30)Jiménez and De Paz 2008; Reynolds et al. 2006peer-reviewed reproduction
Lander (1985)100 × weight ÷ (101.3 − 2.67123 × reps)Jiménez and De Paz 2008; Reynolds et al. 2006peer-reviewed reproduction
Lombardi (1989)weight × reps^0.10Jiménez and De Paz 2008; Reynolds et al. 2006peer-reviewed reproduction
Mayhew et al. (1992)100 × weight ÷ (52.2 + 41.9 × e^(−0.055 × reps))Mayhew et al. 1992 (abstract); Reynolds et al. 2006original abstract
O'Conner et al. (1989)weight × (1 + 0.025 × reps)Jiménez and De Paz 2008; Reynolds et al. 2006peer-reviewed reproduction
Wathen (1994)100 × weight ÷ (48.8 + 53.8 × e^(−0.075 × reps))Jiménez and De Paz 2008peer-reviewed reproduction
Forms as printed in Brzycki 1993, the Mayhew et al. 1992 abstract, Jiménez and De Paz 2008 (Table 1) and Reynolds et al. 2006 (Table 5); re-read 1 October 2026.

The seven equations disagree most at high reps, so the comparison of every published form, its author rep limit and its source settles which one-rep max formula to use.

The spread shows how much the seven published equations disagree for this set; it is not a measured error.

Which equation matters less than how many reps the set had; the next block states what the rep-count study found.

How many reps give the most accurate one-rep max estimate?

A 5-rep set gave the closest estimate among the sets it tested: Reynolds, Gordon and Robergs (2006) compared 5-, 10- and 20-rep sets on the chest press and leg press and found the 5-rep set predicted the one-rep max best, and advised no more than 10 reps in linear equations. Brzycki (1993) limits his equation to sets of fewer than 10 reps.

Brzycki's limit of fewer than 10 reps is his own, stated in the 1993 paper. The Reynolds study tested 70 adults aged 18 to 69, so it says nothing about sets of 3 or 8 reps, or about the squat and deadlift. The calculator follows the same 10-rep line and warns past it: 225 lb lifted 12 times returns 324.11 lb with Brzycki's equation while the seven equations span 288.47 to 324.93 lb (Reynolds et al. 2006).

The validation studies behind that answer are set out by rep count and lift on the accuracy page, which shows how accurate a one-rep max estimate is for a given set.

A rep count also maps to a share of the max; the chart block gives the published shares.

What percentage of a one-rep max matches each rep count?

The NSCA Training Load Chart sets 2 reps at 95% of the one-rep max, 5 reps at 87%, 10 reps at 75% and 12 reps at 70%. It lists 1 to 10 and 12 reps only, with no 11-rep value, and the percentage chart rounds each load to real plates.

The chart is a training-load convention published by the National Strength and Conditioning Association, not a measurement of any lifter, and the calculator's loads tab reads the same rows. For a 100 kg max the 5-rep row gives 87 kg before rounding to plates.

Table 2. Four rows of the NSCA Training Load Chart; the chart lists 1 to 10 and 12 reps and has no 11-rep value.
RepsPercentage of the one-rep max
295%
587%
1075%
1270%
NSCA Training Load Chart, adapted from Landers 1984; the full chart is on the percentage page.

The full chart of loads at all the percentages of a one-rep max, rounded to real plates in kilograms or pounds, is on the percentage page.

The same set behaves differently by lift; the next three blocks say how.

How is a bench press one-rep max calculated?

A bench press one-rep max is estimated from the weight and reps of one bench set with the same seven equations. In LeSuer et al. (1997), five of the seven differed significantly from the measured bench 1RM of 67 untrained college students, so the bench calculator shows every equation's result side by side.

The bench finding comes from the LeSuer et al. abstract, which states how many equations differed, not which ones, so this site prints the count and no ranking.

The bench press one-rep max calculator runs the same set through all seven equations and prints the bench-specific validation beside the result.

The squat block follows the same pattern with its own count.

How is a squat one-rep max calculated?

A squat one-rep max is estimated from one set of back squats with the same seven equations. In LeSuer et al. (1997), six of the seven differed significantly from the measured squat 1RM of untrained college students, which is why the squat calculator reports the spread between equations, not one figure.

The same 1997 study tested the back squat in the same sample of untrained college students; six equations differing is the only squat-specific validation figure this site prints.

The squat one-rep max calculator reports the spread between equations for a back squat set rather than one figure.

The deadlift is the one lift where every equation ran low.

Why does a deadlift one-rep max estimate come out low?

A deadlift one-rep max estimated from reps tends to come out low. LeSuer et al. (1997) found that all seven prediction equations significantly underestimated the measured deadlift 1RM of 67 untrained college students, so the deadlift calculator marks its estimate as a probable underestimate rather than a ceiling.

The direction is stated in the abstract; the size of the underestimate is not, so the deadlift page labels its estimate as probably low without putting a number on how low.

The deadlift one-rep max calculator marks its estimate as a probable underestimate and shows how far the seven equations sit apart.

Before the per-lift rows, the definition block says what the number is.

What is a one-rep max?

A one-rep max (1RM) is the maximum weight a lifter can lift for one repetition of an exercise in weight training; it can also be read as the most force one maximal contraction generates. Coaches set training loads as percentages of it, from 100% at 1 rep to 70% at 12 on the NSCA chart.

The definition follows the Wikipedia lead for one-repetition maximum, which places the term in weight training; a personal record is a lift achieved at any rep count and is not the same quantity. The percentage rows come from the NSCA Training Load Chart. The spread printed beside every estimate here is the disagreement between equations, not the difference from a tested max.

The definition page explains what a one-rep max measures, how it differs from a personal record and an estimated max, and how coaches use it.

The seven equations that estimate it are listed next.

Which equations estimate a one-rep max?

Seven one-rep max equations are in common use: Brzycki, Epley, Lander, Lombardi, Mayhew, O'Conner and Wathen. Each turns the weight and reps of one set into an estimate, and they disagree most at high reps. The formula comparison gives every equation's published form, the lift it was built on and its source.

Mayhew's equation was derived from 184 men and 251 women doing as many bench-press reps as possible in one minute at 55 to 95% of their max (Mayhew et al. 1992). Reynolds et al. describe the linear equations as built on sets of fewer than 10 reps and the nonlinear ones, Lombardi's and Mayhew's, on sets of up to 15 (Reynolds et al. 2006).

Each of the seven equations and their sources, with the lift and rep range it was built on, is set out on the formula comparison page.

The lift changes how the equations behave; the rows below summarise the evidence per lift.

One-rep max by lift

The validation evidence differs by lift. On the bench press five of the seven equations differed from the measured max, on the squat six of seven, and on the deadlift all seven underestimated it (LeSuer et al. 1997). At 70% of the max a leg press set runs to about 19 reps against 14.1 on the bench, in the estimates of a 2024 meta-analysis (Nuzzo et al.).

The three lift rows come from the LeSuer et al. 1997, which reports counts rather than equation names; the leg-press row is a pooled estimate in Nuzzo et al. 2024.

Table 3. What the validation studies found on each lift.
LiftFindingSource
Bench pressFive of the seven equations differed significantly from the measured maxLeSuer et al. 1997, 67 untrained college students
SquatSix of the seven equations differed significantly from the measured maxLeSuer et al. 1997
DeadliftAll seven equations underestimated the measured maxLeSuer et al. 1997
Leg pressAbout 19.0 reps at 70% of the max, against 14.1 on the bench pressNuzzo et al. 2024, meta-analysis estimates
Counts, not equation names: the LeSuer abstract states how many equations differed, not which.

The bench row summarises how the equations performed on the bench press in the one study that tested all seven. The squat row summarises how the equations performed on the squat in the same study. The deadlift row explains why deadlift estimates run low across every equation tested. The leg press row gives the reps a leg press set allows at each percentage compared with the bench press.

Every calculator on the site is listed next.

Every one-rep max calculator on this site

Every calculator on this site runs the same seven equations from the same constants, so a bench, squat or deadlift set gives the same estimate here and on its own page. The calculators for every lift, the formula and chart pages, the effort and testing pages and the scoring pages are listed below as they publish.

A 5-rep max of 100 kg converts to a 112.51 kg one-rep max and an 84.36 kg 10-rep max when Brzycki's equation is run in both directions, and the converter, like every page here, reads the same constants file as this calculator; the seven forms are the set reproduced in Jiménez and De Paz 2008.

The bench, squat and deadlift calculators, the formula comparison and the definition page are linked from their own blocks above; the rest of the site's pages are listed here.

An estimate can also be wrong in predictable ways; the next block lists them.

When a one-rep max estimate is wrong

A one-rep max estimate is predictably off in four cases. Above 10 reps every equation is extrapolating, so the calculator warns (Reynolds 2006). On the deadlift every equation ran low (LeSuer 1997). For a single rep no equation is applied; the weight itself is returned. For a set stopped with reps left, the equations still assume reps to fatigue (Brzycki 1993), so the estimate comes out lower than the lifter's max; the RPE option is for that set.

  • Above 10 reps: every equation is extrapolating, so the calculator warns and prints the estimate with its spread (Reynolds et al. 2006).
  • The deadlift: all seven equations underestimated the measured max in LeSuer et al. 1997, so the estimate is labelled as probably low.
  • A single rep: no equation is applied and the weight itself is returned; no estimate is given above 20 reps.
  • A set stopped with reps left: the equations take reps to fatigue as their input (Brzycki 1993), so the estimate comes out below the lifter's max; rating the set with RPE gives a second estimate for that case.
The four cases in which an estimate is predictably off, with the source of each.

A true one-rep max attempt belongs after a graded warm-up, with safety arms set or a spotter present; pain or injury is a question for a coach or physiotherapist, not a calculator.

The warm-up and attempt protocol the testing page follows is the one the HPRC and NSCA describe: light reps first, load steps of 5 to 10% for upper-body lifts or 10 to 20% for lower-body lifts, and 2 to 4 minutes of rest before attempts.

A true max attempt belongs after a graded warm-up with safety arms set or a spotter present; the protocol for testing a one-rep max directly is on its own page.

Who this page serves is stated in one line per situation.

Lifters this calculator serves

This calculator serves the lifter who has just finished a hard set of up to 10 reps and wants the max without attempting it, the lifter setting today's working weights from a known max, the coach or student who needs the source of each equation, and the lifter who is benching, squatting or deadlifting and wants that lift's own estimate.

A lifter setting today's working weights from a known max needs loads that fit real plates, which the percentage chart rounds for them. The coach or student who needs the source of each equation finds it on the methodology page, and the lifter who wants a bench, squat or deadlift estimate has a calculator for that lift.

The default equation, and the arithmetic the site refuses to headline, are explained next.

The equation this site uses by default, and why

The default equation is Brzycki (1993), the only one verified here against its original paper; its author limit of fewer than 10 reps matches the calculator's cap, and the choice is marked as proposed on the methodology page. It is not the most accurate equation on every lift: in Reynolds et al. (2006) Brzycki's equation underestimated the leg press. The average of the seven equations is not a published method, so it is shown last and labelled arithmetic. A single rep returns the weight itself, where Epley would give 103.33 kg and Mayhew 108.86 kg for 100 kg lifted once.

The Brzycki equation is the only one verified here against its original paper; how constants are verified, and on what date, is recorded on the methodology page. The equation itself is read from Brzycki 1993. In the Reynolds et al. 2006 cross-validation on the leg press, Brzycki's, Lander's and O'Conner's equations all ran low, while Epley's had one of the smallest mean errors, which is why all seven results stay on screen beside the default.

Table 4. Worked example: 100 kg lifted once; in kilograms or pounds the rule is the same.
EquationBare equation resultWhat the calculator returns
Epley (1985)103.33 kg100 kg
Mayhew et al. (1992)108.86 kg100 kg
Why the single-rep rule exists: the bare equations inflate a lift the lifter has already made, so one rep returns the weight itself.

What the calculator does not do closes the page.

What this calculator does not do

This calculator does not store lifts or need an account, does not prescribe a training programme, does not rank the lifter against strength standards built from the lifts other sites collect, does not sell supplements or coaching, and does not estimate a one-rep max above 20 reps. Its optional RPE line reads the Helms 2016 chart and its training-max line uses 90% of the estimate by default, or 85% as a lighter setting; both are labelled.

  • No account and no stored lifts: the calculator runs in the browser, and a result is saved as a link if the lifter wants to keep it.
  • No training programme: the training-max line prints 90% of the estimate, or 85% as a lighter setting, and nothing else.
  • No ranking against strength standards built from the lifts other sites collect; standards appear here only from a dataset the site can cite.
  • No supplements, coaching or equipment rankings for sale.
  • No estimate above 20 reps, and a warning above 10.

The RPE line reads the Helms et al. 2016 chart, where 5 reps at RPE 8 is 81% of the max, an estimated cell built from trained lifters' back squats, on the Zourdos et al. 2016 scale, where 9.5 means no further reps but a heavier load was possible. The training max follows Wendler's original 5/3/1, which calculates from 90% of the max, and his page for beginners, which uses 80 to 85%. Loads round to the plates of the chosen unit, with one pound taken as 0.45359237 kg exactly (NIST) and the kilogram disc set taken from the IPF Technical Rulebook.

The questions lifters ask most follow.

Constants verified 27 September 2026; the writer and the reviewer are named in the byline under the calculator.

who publishes this calculator, how it is funded and what it will not publish is stated on the about page. Rohith Singh reviews the pages that cover practical training, and the reviewer page sets out what the review covers. Errors are corrected under the editorial and correction policy, with a dated entry in the revision log.

One-rep max questions

What is my 1 rep max if I can bench 225 for 12?

About 324.1 lb with Brzycki's equation; the seven equations span 288.5 to 324.9 lb. Twelve reps is past the 10-rep limit, so the calculator warns; a set of 5 predicts closer (Reynolds 2006).

Which 1RM formula is most accurate?

No single equation wins every lift. Reynolds (2006) found a 5-rep set predicted best among the sets tested, and in LeSuer (1997) all seven equations underestimated the deadlift. This site defaults to Brzycki and shows all seven.

Does a one rep max calculator work for bodybuilding sets?

Yes, for a set of 1 to 10 reps taken close to failure. The same equations apply; above 10 reps the estimate carries a warning.

Is there a one rep max calculator app?

No app to install. The calculator runs in any phone browser and stores nothing; a result can be saved as a link.

Should you max out to find your one rep max?

An estimate from a set of 1 to 10 reps avoids a max attempt. When an exact number is needed, for a meet or a test, follow a graded protocol with safety arms set or a spotter present.

What is a training max?

A training max is the share of the one-rep max that Wendler's 5/3/1 programme calculates from: 90% in his original 2009 article, 80 to 85% on his page for beginners. The calculator defaults to 90% and offers 85%; it prints no 5/3/1 week templates.

How do you convert a one rep max from kg to lb?

One pound is 0.45359237 kg exactly. The kg/lb switch converts the entered weight, and loads round to that unit's plates.

Does this calculator compare my lift with other lifters?

No. The site collects no lift data and does not reuse the lifts another site has collected; strength standards appear only from a dataset it can cite.

What does the RPE option in the calculator do?

RPE rates how many reps were left in the set on the Zourdos (2016) scale. With it, a second estimate comes from the Helms (2016) chart, where 5 reps at RPE 8 is 81% of the max, and it is never mixed with the equation estimate.