How this works
A one-rep max (1RM) is the maximum weight you could lift for a single repetition with proper form. It's the foundational number in strength training programming — most periodised programmes prescribe loads as a percentage of 1RM (e.g. 5×5 at 80%, AMRAP at 70%) rather than absolute kilograms or pounds, because it makes the programme scale to whoever's running it. Actually testing your 1RM is risky and exhausting; instead, you do a heavy set of 3-8 reps and use a formula to extrapolate what your 1RM probably is. The accuracy is good enough for programming for everyone except elite competition lifters.
Four formulas dominate. Epley (1RM = w × (1 + reps/30)) is the simplest and most-cited; it's very accurate up to about 10 reps and starts to over-estimate beyond that. Brzycki (1RM = w × 36 / (37 − reps)) is similar in spirit but slightly more conservative for higher reps; many strength coaches prefer it for that reason. Lombardi (1RM = w × reps^0.10) is a power-law fit that handles the full 1-30 rep range smoothly. There's no "right" answer — they're all empirical curves fitted to lifter data, and they tend to agree to within 5% in the 3-8 rep range that matters for programming. The calculator above shows all three and you can pick whichever you trust most.
Two practical points. (1) Test reps in the 3-8 range for the most accurate prediction. Below 3 reps you don't need a formula — you're close enough to a true 1RM. Above 8 reps the relationship between strength and endurance starts to dominate and the prediction degrades fast (a 20-rep set tells you more about your conditioning than your 1RM). (2) The percentage-of-1RM table is the practical output for programming. 90-95% of 1RM is typically 2-3 rep territory; 80-85% is 5-6 rep working sets; 70-75% is hypertrophy 8-10 rep range; 60-65% is volume 12-15 rep work. Different programmes use different load ranges for different goals — strength coaches like Wendler 5/3/1 use 65-95%, hypertrophy programmes like Boring But Big use 50-70%, powerlifting peaking blocks use 90%+. The right percentage depends on what you're training for.
The formula
w is the weight lifted (kg or lb — both work; the formulas are unit-agnostic). reps is the number of clean repetitions completed at that weight. Brzycki breaks down at 37 reps (denominator → 0); restrict input to ≤ 15 reps for any formula to keep estimates reliable. The training-percentage table multiplies the chosen 1RM by 50% through 95% in 5% steps so you can read off working weights for any rep range.
Example calculation
- 5 reps at 100 kg / 220 lb on the bench press.
- Epley: 100 × (1 + 5/30) = 100 × 1.167 = 117 kg / 257 lb 1RM.
- Brzycki: 100 × 36 / 32 = 112.5 kg / 248 lb. Lombardi: 100 × 50.10 = 117 kg / 257 lb.
- For 5×5 working sets at 80% of 117 kg = 94 kg / 207 lb.
Frequently asked questions
Which formula should I use?
For most lifters, Epley is fine — it's the most cited, the simplest, and accurate enough in the 3-8 rep range that matters for programming. Brzycki is slightly more conservative for higher reps (8+); pick it if you'd rather under-estimate your 1RM than over-estimate (better safe than injured). Lombardi handles the full 1-30 rep range smoothly and is worth checking against Epley if you tested with high reps. The honest answer is that all three formulas are within ~5% of each other in the rep range that matters, and the difference is smaller than the day-to-day variation in your real strength. Pick one, use it consistently, and adjust based on what your training tells you.
How many reps should I test with?
Aim for 3-8 reps. Below 3 you don't need a formula — you're close enough to actual 1RM that you can just call it that. Above 8, the prediction degrades because muscular endurance starts to overshadow strength as the limiting factor. The sweet spot for accuracy is 5 reps — enough to be safe (you're not maximally loaded) but not so many that endurance dominates. For working out a 1RM from a recent training session, just use whatever set you actually did rather than testing — if you logged 6×185 lb in your training journal, plug that in. Real-life sets give you a more accurate prediction than a contrived test because they reflect normal performance, not adrenaline-fueled effort.
Why do my predicted 1RMs vary between exercises?
Because the formulas were fit to general lifter data and individual lifts have different "rep economies". Most lifters find their bench press follows Epley/Brzycki closely; their squat over-predicts (you can usually do more squat reps at a given % than the formula expects); their deadlift under-predicts (you usually can do fewer deadlift reps at a given % because deadlifts crush the central nervous system). The fix is to track your actual 1RMs over time and learn your personal rep-to-1RM curves for each lift. After 6-12 months of consistent training and logging, you'll know that "5 reps at 90% of squat 1RM" is achievable for you while "5 reps at 90% of deadlift 1RM" might not be. Use formulas as starting estimates, not gospel.