How to Estimate a One-Rep Max Without Testing One
A one-rep max is the heaviest weight you can lift once. It is also the number most training programmes are written around, which creates an awkward problem: to follow the programme you need the number, and to get the number you have to attempt a maximal single — the most fatiguing, most technically demanding, highest-risk thing you can do in a gym, and the thing a beginner is least equipped to do safely.
The way round it is to estimate the max from a set you have already completed. Lift a weight you can handle for a handful of reps, count them honestly, and a formula converts that into a projected single.
The two formulas worth knowing
Two estimates dominate, and both are simple enough to do on paper:
- Epley: 1RM = weight × (1 + reps ÷ 30)
- Brzycki: 1RM = weight × 36 ÷ (37 − reps)
Take a set of 100kg for 5 reps. Epley projects 100 × (1 + 5/30) = 116.67kg. Brzycki projects 100 × 36/32 = 112.5kg. The One-Rep Max Calculator returns both plus their average, 114.58kg, because neither formula is authoritative and the spread between them is useful information in itself.
At one rep both formulas return the weight unchanged, which is not a mathematical accident so much as a necessary correction: Epley's raw form gives 96.8% of the weight at a single rep, which is nonsense, so a single is treated as the max by definition.
Where the estimate is trustworthy, and where it is not
Accuracy falls off as reps climb, and it does so for a physiological reason rather than a mathematical one. A set of three is limited by how much force you can produce. A set of fifteen is limited by fatigue tolerance, breathing, and how much discomfort you are willing to sit in — qualities that vary enormously between lifters and have only a loose relationship with maximal strength.
Two people who both squat 100kg for 15 reps may have genuinely different singles. Two people who both squat 100kg for 3 reps almost certainly do not. That is why the calculator caps input at 12 reps: past that the formulas are extrapolating from the wrong quality.
Practically, treat a set of 3–6 reps as the sweet spot. Close enough to maximal that strength is the limiter, far enough from a true single that the set is safe to perform and repeatable.
Honest reps or nothing
The estimate assumes the set was taken close to failure. If you stop a set of 8 with four comfortable reps left, the formula still treats it as a hard 8 and hands you a projection that is meaningfully low. If you count a rep that was cut short, or one where a partner took some of the weight, the projection is high — which is the more dangerous error, because the next session's working weights are built on it.
This is the single largest source of error in practice, and it is not a limitation of the formula. It is a limitation of self-report. If your estimate keeps producing working weights that feel wrong, the rep counts are usually the reason before the maths is.
Turning the estimate into working weights
An estimated max is only useful once it becomes a weight on the bar. Feed it into the Rep-Max Percentage Calculator to get the loading for each rep target, or read the worked Rep-Max Plate Chart, which goes one step further and shows the plates for each side — including the rows where the prescribed weight is not a weight a barbell can actually be loaded to.
That last detail catches people out. A 140kg estimated max at 90.91% asks for 127.27kg. There is no combination of plates that makes 127.27kg. You load 127.5kg and get on with it.
How often to re-estimate
Re-estimate when the working weights stop feeling right, not on a calendar. If a prescribed set of five is moving faster than it should for weeks on end, the max the programme is built on has been outgrown. If the same set is grinding, either the estimate was optimistic or something outside the gym — sleep, food, stress — is the actual limiter, and adjusting the number will not fix that.
A practical rhythm for most lifters: take a hard set of three to five on your main lift every four to six weeks, run it through the calculator, and let that steer the next block. You will never need to test a true single unless you intend to compete, and even then the testing belongs in a peaking block, not in general training.
What the number is and is not
An estimated max is a planning figure. It tells you roughly where your working weights should sit and lets you compare this month against last month on a consistent scale. It is not a promise that the weight would move on the day, and it is not a target to chase for its own sake.
The lifters who get the most out of these formulas treat them as a thermostat rather than a trophy: a number that quietly sets the temperature of the next block, gets updated when it drifts, and is otherwise not thought about much.
Which lift to estimate from
The formulas were built around barbell lifts and behave best there. A few practical distinctions:
- Squat and bench press are the most reliable. Both are stable, both fatigue predictably across a set, and both have a clear point where the rep genuinely fails.
- Deadlift is less reliable at higher reps. Grip often fails before the legs and back do, and a set that ended because the bar slipped tells you about your grip rather than your pull. If you use straps for the estimate but not in training, the number describes a lift you do not actually do.
- Overhead press works well, but the small absolute weights magnify small errors. Two kilos of measurement noise is over four percent of a 45kg press.
- Machines and dumbbells should be treated with scepticism. A machine's stack is not a true load because of leverage and cable ratios, and dumbbell work is limited by stabilisation as much as by force. Track the weight and reps directly instead of converting them into an imaginary single.
When the estimate is obviously wrong
Sometimes the number that comes back does not match your experience of the lift. Before deciding the formula is broken, check the usual causes in order.
The set was not close to failure. This is the most common by a distance. A set that stopped when it got uncomfortable rather than when the next rep would not have moved produces an estimate that is too low, sometimes by ten percent or more.
The rep count was generous. A partial rep, a bounced rep, a rep a spotter helped with. Each one inflates the projection, and the inflation lands directly on next week's working weights.
You picked the wrong rep range. Estimating from a set of twelve when you normally train threes will overstate your single if you are an endurance-leaning lifter and understate it if you are not. Redo it from a set of four or five.
You are comparing across conditions. An estimate from a fresh first exercise and one from the fourth movement of a long session are not the same measurement. Keep the conditions consistent if you want the numbers to be comparable.
Using the spread as information
The calculator returns Epley, Brzycki and their average together, and the gap between the two is worth a glance rather than being ignored. In the low rep ranges they sit close — a few kilos apart on a mid-weight lift. If you are seeing a large gap, you are estimating from too many reps, and the honest fix is a heavier set next time rather than picking whichever number you prefer.
When they agree closely, you can be reasonably confident in the figure. When they do not, the disagreement is telling you that the input was outside the range where either formula is dependable.