Quick answer
Your one-rep max (1RM) is the heaviest weight you can lift once. It can be estimated from a set of several reps with a formula such as Epley, weight × (1 + reps ÷ 30), instead of being tested. For 80 kg lifted for 8 reps, the calculator's four formulas give 96.0 kg (O'Conner) to 101.3 kg (Epley), with Brzycki and Lombardi in between, and average about 98.8 kg.
What a one-rep max actually is
Your one-rep max (1RM) is the heaviest weight you can lift for a single complete repetition with correct technique. It's the reference point strength programs are built around: instead of writing "lift heavy today," a program can say "work at 80% of your 1RM for 5 sets of 5," and that number means the same relative effort whether you bench 60 kg or 200 kg. A 2003 master's thesis that tested prediction equations of this kind (Schroeder; see the audit trail) notes that attempting a 1RM takes a large amount of concentration and mental preparation and may involve some risk of injury, while adding that injury rates for 1RM tests have not been reported. It says a way to predict a 1RM without the lift is appealing for athletes who are unwilling to attempt one or have a higher potential risk of injury, and that a number of equations have been developed as a result to estimate a 1RM from a submaximal set (one you can do for several reps, not just one).
An estimate compared with a tested max
A tested 1RM is a measurement. HPRC (Human Performance Resources by CHAMP, at the Uniformed Services University) describes it as the heaviest weight you can lift just once, found by working up through warm-up sets to maximal attempts (the protocol is in the FAQ below). An estimate is a calculation from a set of several reps. HPRC calls a multiple-repetition test a good option for someone lifting alone with no one to spot them, and says such tests are not very reliable above a 5-rep max for power exercises, because form breaks down. The estimate has its own error: in two of the studies the thesis summarizes, the average prediction missed the tested bench press 1RM by 2.2 to 5.4 lb (five equations, in untrained students) and by 4.80 to 14.2 kg (three equations, in college football players), and a third cautions that large errors for individuals are possible (see the FAQ). The thesis also reports a finding that the relationship between submaximal lifting and 1RM strength changes as a person adapts to a training program. We have not read those studies themselves.
The four formulas this calculator uses
All four formulas take the weight you lifted and the reps you got and return a higher weight. They disagree because each is a different curve: Epley's and O'Conner's are straight lines in reps, Brzycki's has reps in a denominator, and Lombardi's raises reps to the power 0.10.
The publication cited for each equation in the reference list of a 2003 University of Hawai'i master's thesis (Schroeder), which prints all four in a table it describes as derived from LeSuer and colleagues (1997) and Mayhew and colleagues (1999): Epley, a 1985 poundage chart; Brzycki, a 1993 article in the Journal of Physical Education, Recreation and Dance; Lombardi, the 1989 book Beginning Weight Training; O'Conner, Simmons and O'Shea, the 1989 book Weight Training Today. We could not open any of the four, so this page cannot say what data, if any, each was fitted to. The thesis prints Epley as 1 + 0.0333 × reps and Brzycki as 100 ÷ (102.78 − 2.78 × reps); this page uses reps ÷ 30 and 36 ÷ (37 − reps), which differ from the printed forms by at most 0.09% up to 20 reps.
Epley and O'Conner are linear — they add a fixed percentage of the lifted weight per rep, so they scale smoothly: Epley's 3.3% a rep keeps it above the four-formula average, and O'Conner's 2.5% keeps it below the average from 2 reps up. Brzycki and Lombardi are curved, in opposite directions: Lombardi flattens and becomes the lowest estimate from 11 reps, while Brzycki steepens and becomes the highest. Notice that at exactly 1 rep, Brzycki and Lombardi return the weight you lifted unchanged, while Epley and O'Conner still add a small bump — a known quirk of the linear formulas, not a bug.
How the four formulas compare
Epley runs a few percent above Brzycki at low reps, the gap closes by 10 reps, and past 10 Brzycki climbs above Epley. Lombardi gives the highest estimate from 2 to 5 reps before flattening out, and O'Conner, adding only 2.5% per rep, is the lowest of the four from 6 to 10 reps. Past 10 reps the four pull apart quickly: Brzycki climbs to the top, Lombardi drops to the bottom with O'Conner just above it, and they are about 9% apart at 11 reps and 12% at 12. Which of them comes closest to a tested max is not settled by the study summaries we read (see the FAQ), so the calculator shows all four instead of picking a winner and leads with their range; the average is shown alongside as one summary figure.
Worked example 1 — bench press
You bench 80 kg for 8 reps to fatigue.
Brzycki: 80 × 36/(37−8) = 99.3 kg
Lombardi: 80 × 8^0.10 = 98.5 kg
O'Conner: 80 × (1 + 0.025×8) = 96.0 kg
Average 1RM ≈ 98.8 kg
Worked example 2 — back squat
You squat 140 kg for 5 reps to fatigue.
Brzycki: 140 × 36/(37−5) = 157.5 kg
Lombardi: 140 × 5^0.10 = 164.4 kg
O'Conner: 140 × (1 + 0.025×5) = 157.5 kg
Average 1RM ≈ 160.7 kg
Notice the spread is proportionally tighter than in the 8-rep bench example above (about 4% of the estimate versus 5%), and the four stay within about 7% of each other up to 10 reps before spreading quickly.
Reading the percentage table: training zones
Once you have a 1RM estimate, a percentage of it is the number a program usually prescribes. The table above takes its percentages and repetition counts from the National Strength and Conditioning Association's training load chart (adapted from Landers, 1984), which pairs each percentage of 1RM with the most repetitions that can be done at it. The chart ends at 70% (12 reps), so the table does too. The zone labels apply the cut-offs that the Uniformed Services University's Human Performance Resources by CHAMP (HPRC, 5 July 2022) prints when it reproduces the NSCA's recommended loads: muscular strength, 4 to 6 reps at 85% of 1RM or more; hypertrophy (muscle growth), 6 to 12 reps at 67% to 85%; muscular endurance, at least 12 reps at less than 67%. The labels use the percentage cut-offs only; at exactly 85% the two ranges meet and the table lists it as strength, and the endurance zone starts below the lowest row. Other sources draw the lines elsewhere: a 2021 review by Schoenfeld and colleagues describes the "repetition continuum" as strength at 1 to 5 reps with 80% to 100% of 1RM, hypertrophy at 8 to 12 reps with 60% to 80%, and endurance at 15 or more reps below 60%, and argues that muscular adaptations can be obtained across a wide spectrum of loading zones. The zones are one convention among several.
Not medical or coaching advice
This tool does strength math. It is not medical, coaching or physical-therapy advice, and it cannot account for injury history, joint health or technique quality. If you test a true 1RM instead of estimating one, the protocol described in the FAQ below is the NSCA's, as summarized by HPRC.
FAQ
How many reps should I use to get the most accurate 1RM estimate?
The study summaries we read do not point to a best number of reps. Two studies summarized in a 2003 thesis suggest predictions are closer from sets of 10 reps or fewer than from longer sets (Mayhew and colleagues, 1995, for six equations; Chapman and colleagues, 1998, for a test with a fixed 225 lb load), and the thesis reports that Brzycki held his formula valid only for fewer than 10 reps. We have not read those studies themselves. Past 10 reps the four formulas on this page also spread apart quickly: at 12 reps the highest is 12.3% above the lowest.
Why do the four formulas give different numbers for the same set?
Each one is a different curve. Epley and O'Conner are both straight lines that add a fixed percentage per rep, but not the same size: Epley's runs a little high, sitting above the four-formula average at every rep count from 2 to 12, while O'Conner's smaller bump does the opposite: it sits below that average from 2 reps up and is the lowest of the four from about 6 to 10 reps. Lombardi's curve flattens at higher reps, making it the most conservative estimate there; Brzycki's curve runs the other way: conservative at low rep counts, but it climbs faster than the others past about 10 reps and becomes the highest of the four. At 2 reps the highest of the four is 4.2% above the lowest and at 3 reps 5.4% above it; by 10 to 12 reps the gap grows from 6.7% to 12.3%. We could not open the original publications (a 1985 poundage chart, a 1993 journal article and two 1989 weight-training books), so we cannot say what data, if any, each curve was fitted to.
Is an estimated 1RM as accurate as actually testing my max?
No. It is an estimate, not a measurement. In two of the studies a 2003 thesis summarizes, the average prediction missed the tested bench press 1RM by 2.2 to 5.4 lb for five equations (LeSuer and colleagues, 1997) and by 4.80 to 14.2 kg for the Epley, Lander and Brzycki equations (Ware and colleagues, 1995), and a third (Mayhew and colleagues, 1995) cautions that large errors for individuals are possible. We have not read those studies themselves. The same thesis notes that a 1RM attempt may involve some risk of injury, while adding that injury rates for 1RM tests have not been reported.
Which of the four formulas should I trust most?
None of them is shown to be best. The studies summarized in a 2003 thesis disagree: one (Mayhew and colleagues, 1995; 220 men) found the Brzycki equation appeared most accurate under 10 repetitions, another (LeSuer and colleagues, 1997; 67 untrained students) found that five of the seven equations it tested, Brzycki, Epley, Lombardi and O'Conner among them, underestimated the bench press 1RM by 2.2 to 5.4 lb on average, and a third (Ware and colleagues, 1995; 45 football players) found Epley and Brzycki overestimated it, by 4.80 kg and 14.2 kg on average. We have not read those studies themselves. Past 10 reps the four spread apart quickly, with Brzycki climbing highest and Lombardi falling lowest. This calculator shows all four plus their average so that it does not rest on one formula.
What does testing a true 1-rep max involve?
It is a single maximal effort. The protocol the NSCA recommends, as summarized by the Uniformed Services University's HPRC, builds up to it: a light warm-up set of 5 to 10 reps, a heavier set of 3 to 5 reps, then a set of 2 to 3 reps near the maximum, with rests of 1 to 4 minutes between steps, followed by maximal attempts that are raised or lowered until one repetition can be completed with proper technique. HPRC also notes that a multiple-repetition test is an option for someone lifting alone with no spotter. This page does not assess injury risk and is not medical or coaching advice.