Free tool
Enter the weight you lifted and how many reps you got with it. This 1RM calculator runs five established formulas, shows you all five, and gives you the average — plus the training percentages to program from.
Reps are capped at 12. Past that these formulas stop describing strength and start describing work capacity, and Brzycki divides by (37 − reps), which collapses entirely at 37.
Estimated one rep max
115.1kg
Mean of five formulas. They range from 112.5 to 117.5 kg — 4.3% apart.
Two to five reps is where these formulas were validated and where they agree most closely. This estimate is in that range.
This page runs five 1RM equations that have been published for decades — Epley, Brzycki, Lombardi, O'Conner and Wathan — and averages them. They are written out in full further down, and anyone can look them up in a textbook. There is nothing proprietary here: every visitor gets the same arithmetic applied to the same two numbers.
One set is also thin evidence. These equations were fitted to groups of lifters rather than to you, they assume a fixed relationship between reps and percentage of maximum that no two people share, and a single set on a single day knows nothing about how you slept or how far into a hard week you are. Any estimate built from one set is inherently rough — which is why the spread between the five is shown rather than hidden. That spread is the honest width of the answer.
FORGE REPS does not work the way this page works. So treat this as a free public calculator you can run in ten seconds and program from — a ballpark number, not a personalised recommendation, and not medical advice.
Every row below is a percentage of the estimated max, rounded to a load you can actually build on a bar. The rep counts are what a trained lifter typically manages at that percentage — a description, not a target. If you beat them comfortably, your real max is higher than this estimate.
| %1RM | Load (kg) | Typical reps | What it trains |
|---|---|---|---|
| 95% | 110 | 2–3 | Peak strength |
| 90% | 102.5 | 3–4 | Strength |
| 85% | 97.5 | 5–6 | Strength |
| 80% | 92.5 | 7–8 | Strength and size |
| 75% | 87.5 | 9–10 | Size |
| 70% | 80 | 11–12 | Size and volume |
| 65% | 75 | 13–15 | Volume, technique |
| 60% | 70 | 16–20 | Warm-ups, speed work |
Loads are rounded to the nearest 2.5 kg, since that is the smallest jump most gyms can make with a pair of plates.
Your one rep max is the heaviest weight you can move for a single clean rep. It matters because almost every serious program is priced against it: five sets of five at 80%, a top single at 92%, back-off work at 65%. Those instructions need a number to take a percentage of, and going into the gym to find that number the hard way costs you a session and carries the highest injury risk of any rep you will ever do.
An estimate solves that. You take a set you were going to do anyway, push it close to failure with clean technique, and read the max off the arithmetic. It is not as accurate as a tested max — nothing is — but it is accurate enough to program with, and you can refresh it as often as you train instead of twice a year.
All five formulas take the same two inputs: w, the weight you lifted, and r, the reps you completed with it. At one rep every one of them returns the weight itself, which is the only honest answer — a single is a measured max, not an estimate.
1RM = w × (1 + r ÷ 30)
1RM = w × 36 ÷ (37 − r)
1RM = w × r^0.10
1RM = w × (1 + r ÷ 40)
1RM = w × 100 ÷ (48.8 + 53.8 × e^(−0.075 × r))
Each equation was fitted to a different sample of lifters doing different exercises. They all assume a fixed relationship between reps and percentage of maximum, and no two lifters share that relationship. Someone with high rep endurance grinds out eight at 80% and gets underestimated; someone explosive but fatigue-prone manages five and gets overestimated.
At two to five reps the five estimates usually land within about five percent of each other, which is tight enough to be useful. By twelve reps they spread past eleven percent. Most calculators pick one formula and print a single confident number; that number is no more accurate, it just hides the uncertainty. Showing the spread tells you exactly how much of your estimate is maths and how much is guesswork.
Accuracy degrades as reps rise. Above roughly ten reps, what you are really measuring is work capacity, not maximal strength. Treat anything from a high-rep set as a rough ballpark and re-test with a heavier, shorter set when you need a number you can trust.
A genuine maximal attempt is a maximal neural effort. It taxes your central nervous system and your connective tissue far more than the weight on the bar suggests, and it can take several days before your working sets feel normal again. It is also the rep where form breaks down under load, which is precisely when people get hurt.
Two or three tested maxes a year is plenty for most lifters, usually at the end of a training block when you are peaked and fresh. Between those, estimate from a hard set of three to five. You get a number accurate enough to program with, at a fraction of the cost, from a set that was already in your program.
Pick the row that matches the job. Strength work lives at 85–95% for doubles, triples and heavy fives. Most hypertrophy work sits at 65–80%, where you can accumulate real volume without every set being an event. Warm-ups and speed work belong at 55–70%.
Progressive overload is what turns that table into progress, and the usual way of handling it is not something this page invented. Standard double-progression practice is to work at a given percentage until the prescribed reps feel comfortable rather than close to failure, then recalculate the estimated max from a recent best set and rebuild the table around the new number; most published templates suggest refreshing it every four to six weeks. Every load moves up together, which is the point of pricing work against a max at all — overload, as those templates describe it, is not adding weight whenever you feel good, it is a systematic shift of the whole scale once you have earned it. That is a general rule of thumb from the training literature, not a rule this page or the FORGE REPS app applies to you.
Two cautions. These estimates are least reliable on deadlifts, where most people manage fewer reps at a given percentage than the formulas predict, and on isolation work, which was never what they were built for. And they describe a fresh, well-rested lift — the same set on the back end of a hard week will read low.
None of them wins across every lifter and every lift. Brzycki reads low on short sets and high on long ones, since its denominator runs away as the reps climb; Lombardi and O'Conner stay the most conservative at high reps, with Epley and Wathan between them. Which one matches you depends on your own rep endurance, which is why this page shows all five and takes the mean — an average of several validated equations is more robust than any one of them, and the spread tells you how much to trust it.
Two or three times a year for most lifters, at the end of a block. A maximal attempt is expensive to recover from and carries the highest injury risk of anything you do in the gym. Estimate the rest of the time.
It works best on the big compound barbell lifts — squat, bench press, deadlift, overhead press — which is what these formulas were validated on. On curls, lateral raises and machines the estimate drifts, because small muscle groups fatigue on a different curve.
Two to five, taken close to failure with clean technique. That is the range these equations were built on and where they agree most closely.
This calculator is a standalone free tool, not a preview of the app. FORGE REPS is a separate product, for people who would rather have their training tracked properly than carried around in their head. It launches in the browser first — no install. Join the waiting list and you'll hear the day it opens.
Check your inbox