A single completed set is enough to estimate your maximum load for one repetition, and more usefully to derive the loads you should be training at. That second use is the one that matters: nobody trains at their max, everybody trains at a percentage of it.
The calculator below shows both reference formulas side by side instead of one number, because they don't say quite the same thing and the gap between them is information.
Estimated 1RM—
| Formula | Result |
|---|
Your working loads
| % of 1RM | Load | Reps |
|---|
The two formulas, and why they differ
Epley multiplies your load by one plus the reps divided by thirty. Brzycki multiplies it by thirty-six divided by thirty-seven minus the reps. Two descriptions of the same curve, fitted on different populations.
They agree exactly at ten repetitions. Below that, Brzycki returns the more cautious number. Above it, the reverse. The gap sits around three to four percent on a set of five, which is smaller than the day-to-day variation in the same person.
Put another way: the precision of the formula isn't what limits the result. The set you feed it is.
What the calculator cannot know
Three things, and they weigh more than the choice of formula.
The first is whether your set was genuinely close to failure. An estimate made from five comfortable repetitions understates the max, sometimes by ten percent. The calculation assumes you couldn't have managed a sixth, or nearly. If you left two in reserve, tell the calculator by adding those two to the count.
The second is the movement. A bench press estimated from a set of five is reliable. A deadlift is slightly less so, because technique degrades quickly as the load climbs. And a movement where balance gives out before the muscle does, like a loaded lunge, suits the exercise poorly.
The third is the rep range. Beyond twelve, the estimate comes apart: what limits the set is no longer maximal strength but muscular endurance, and those aren't the same quality. The calculator refuses to answer past that threshold rather than produce a number that merely looks tidy.
The working loads, which are the real point
The percentage table is what you'll use every week. The matching rep ranges are a working convention, not a law.
At 85 percent and above you're training pure strength: few repetitions, plenty of rest, technique that doesn't move. Between 70 and 80 percent sits the range where muscle gets built most efficiently, and that's where most people's sessions live. Below 65 percent you're training strength endurance and technique, which is useful without being the main lever.
One reference point beats the percentage on its own: the last two or three repetitions of a set should genuinely cost something while still looking like the first ones. If rep twelve looks like rep three, the load is a warm-up whatever the arithmetic says.
Reference points, and what they're worth
| Level | Bench press | Squat | Deadlift |
|---|---|---|---|
| Beginner after six months | 0.75 × body weight | 1.0 × | 1.25 × |
| Intermediate, two years | 1.0 × | 1.5 × | 1.75 × |
| Advanced | 1.25 × | 1.75 × | 2.25 × |
These multiples are orders of magnitude for an adult man, and they shift noticeably with build, sex and training history. They're for placing yourself, not for judging yourself. Someone with long arms will always have a bench press lagging their squat, and it says nothing about their training.
Should you test your actual max
Rarely, and this is where common practice diverges from what's useful.
A one-rep test costs a week of recovery, needs a spotter and assumes flawless technique under maximal load. It delivers an exact number that the calculator gets within three percent of from a set of five, which costs a normal session.
The real test keeps two uses. A competition, where the number is the point. And a calibration every six to twelve months, if you fancy it. For steering training week to week, a set of five read by the calculator is plenty, and it has the advantage of being repeatable often.
How the number ages
An estimated 1RM isn't a property, it's a photograph. It moves with sleep, stress, its position in the week, where you are in a training block. A five percent swing from one day to the next is ordinary in the same person, and it signals no loss of strength.
Hence a habit worth more than precision: re-estimate every four to six weeks, on the same movement and under the same conditions, and follow the trend. Three rising estimates say something. One isolated estimate says almost nothing.
FAQ
Is an estimated 1RM reliable?
Within three or four percent when the set runs between three and eight repetitions and was close to failure. That's tighter than natural day-to-day variation, so it's more than enough for choosing working loads.
How many repetitions give the best estimate?
Between three and five. Below that you're approaching a real test with the risks attached. Above eight, muscular endurance starts to weigh in and the estimate drifts. Five repetitions is the usual compromise.
Epley or Brzycki?
Both, and look at the gap. They coincide at ten repetitions, Brzycki is more cautious below, Epley more cautious above. Taking the average of the two is a reasonable call, and it's what the calculator keeps as the working value.
Why does the estimate stop working beyond twelve repetitions?
Because what limits the set changes nature. At fifteen or twenty repetitions it's the ability to sustain an effort that gives out before maximal strength does, and the two qualities aren't linked closely enough for one to predict the other.
Do I need a separate 1RM per exercise?
Yes, and it matters. Strength is specific to the movement: a percentage calculated from a squat has no validity on an overhead press. Every main movement deserves its own estimate.
How do I know my set was genuinely close to failure?
By the speed of the last repetition. If it comes out at the same rhythm as the first, at least two or three were left. If it slows noticeably without technique breaking, you were one or two from failure, which is the ideal state for estimating.
Is a 1RM useful if I don't do powerlifting?
It's useful for choosing loads, which anyone lifting anything with intent needs. Without a reference point, most people train too light for months without knowing it, and that's the most ordinary cause of stalling in a gym.
How often should I re-estimate?
Every four to six weeks, under the same conditions, following the trend instead of the number of the day. Re-estimating every session only measures noise.
Does it work for bodyweight movements?
Only by adding the load you actually move. For a pull-up, feed it your body weight plus any belt weight, and read the result the same way. The percentages then translate into assistance or added load rather than plates, which is the part people find confusing: 80 percent of a weighted pull-up max often means removing weight, not adding it.
Should women use the same formulas?
The formulas themselves hold: they describe a load-repetition curve, not a physiology. What differs is the reference table above, which is built on adult men and understates nothing so much as it simply doesn't apply. Use the calculator, ignore the multiples, and compare yourself with your own estimate from six weeks ago.
Can I enter pounds?
Yes, the kg or lb switch at the top of the calculator handles it. Epley and Brzycki are multiplicative, so a load entered in pounds returns a 1RM in pounds: there's no conversion to get wrong, only the label changes. The body-weight multiples in the reference table are ratios, so they hold in either unit.
A number doesn't build a week
Knowing your loads helps, and it says nothing about how many sets to run, how often to return to a movement, or when to back off. The Vitality Archetype test measures your profile across 5 dimensions in 27 questions and 3 minutes, then generates a week where strength work is dosed for your recovery, with a progression that follows your session feedback. The percentage then becomes a session, placed on the right day.