Direct answer
A one-rep max calculator turns a completed load-and-repetition set into an estimate of the load that might correspond to one repetition. Different equations encode different load–repetition relationships, so the honest output is a comparison rather than a guaranteed maximum. Record the exercise, load unit, repetition count, technique and effort context; compare like with like; and never treat an estimated 1RM as a tested result, a safe attempt, or an automatic training prescription.
Visual explanation
Equation agreement changes with the repetition count
What this calculation tells you
The calculation answers a narrow question: given one entered load and a completed repetition count, what values do three named repetition-based equations produce? It can help standardize a training log or compare repeated same-exercise records when direct testing is not the task.
It cannot observe range of motion, repetition quality, assistance, proximity to failure, equipment differences, fatigue, pain, readiness, or whether the set followed a recognized test protocol. Those missing facts limit interpretation even when the arithmetic is exact.
Where it is used
Strength logs
Convert repeatable same-exercise sets into clearly labelled estimates so changes can be reviewed without calling them tested maximums.
Coaching analysis
Compare formula spread and set context before deciding whether an apparent change is meaningful enough to investigate.
Sport science records
Retain the equation, exercise and testing conditions so an estimated value remains auditable rather than becoming an unlabeled number.
Equipment planning
Use an estimate as one input to a separate plate-loading or scenario calculation after a load has been deliberately selected.
When this guide helps
- You completed five repetitions at a known load and want to compare three common estimates.
- Two training-log entries use the same exercise and repetition count but different loads.
- A single-number app result needs checking against the individual equations behind it.
- A higher-repetition set produces a wider equation range and you need to communicate that uncertainty.
What the three equations are actually estimating
A tested 1RM is the greatest load completed for one repetition under a defined exercise and technique standard. A repetition-based estimate is different: it takes a submaximal set and extrapolates beyond the observed repetitions. The calculator therefore labels Epley, Brzycki and Lombardi separately and reports their mean as a summary of those displayed models—not as a fourth validated test.[2][4]
All three equations use w for the completed load and r for the completed repetition count. Because the load appears as a multiplier, the unit carries through unchanged: kilograms in gives kilograms out, pounds in gives pounds out. Mixing units between records creates a false comparison even though each individual calculation still returns a number.
Research comparing multiple-repetition tests with measured 1RM shows that prediction depends on the exercise and repetition range. In one bench-press and leg-press study, 5RM data predicted the measured result more accurately than 10RM or 20RM data within that sample. That finding supports caution with longer extrapolations; it does not establish a universal best equation for every lift or person.[1]
| Equation | Implemented relationship | What changes the estimate |
|---|---|---|
| Epley | w × (1 + r ÷ 30) | Adds one-thirtieth of the load per repetition |
| Brzycki | w × 36 ÷ (37 − r) | Uses a shrinking denominator as repetitions increase |
| Lombardi | w × r^0.10 | Uses a power relationship with repetition count |
Worked example: 100 load units for five repetitions
Enter a completed load of 100 and five repetitions. Epley gives 100 × (1 + 5 ÷ 30) = 116.67. Brzycki gives 100 × 36 ÷ (37 − 5) = 112.50. Lombardi gives 100 × 5^0.10 = 117.46. The calculator mean is (116.67 + 112.50 + 117.46) ÷ 3 = 115.54 load units.
The individual estimates span 4.96 load units from lowest to highest. That range is not a statistical confidence interval and the mean is not guaranteed to be closer to a measured maximum. It is simply a transparent way to show what the three implemented equations say about the same entered set.
| Output | Substitution | Estimate |
|---|---|---|
| Epley | 100 × (1 + 5 ÷ 30) | 116.67 |
| Brzycki | 100 × 36 ÷ 32 | 112.50 |
| Lombardi | 100 × 5^0.10 | 117.46 |
| Mean of displayed estimates | (116.67 + 112.50 + 117.46) ÷ 3 | 115.54 |
Reproduce this worked caseOpen Enter 100 and 5 to reproduce the three estimates and their displayed mean.
Changed-input comparison: the same load for more repetitions
Now keep the load at 100 and change the repetition count from five to ten. The Epley and Brzycki estimates both become 133.33, while Lombardi becomes 125.89. Their mean is 130.85 and their lowest-to-highest spread is 7.44. The set contains more observed repetitions, yet the equations must extrapolate a different part of the load–repetition relationship and no longer agree as closely.
The comparison does not prove that the ten-repetition set is better or worse evidence for a specific athlete. It shows sensitivity to one input and to equation choice. A large change in estimated 1RM can reflect extra repetitions, a different effort level, changed technique, or a genuinely stronger performance; the output alone cannot separate those explanations.
A large meta-regression using data from roughly 7,000 people found meaningful variation in the repetitions completed at percentages of 1RM and exercise-specific differences, including between bench press and leg press. That is why a universal repetitions-to-percentage table—or one equation treated as exact—can hide real uncertainty.[3]
| Set | Epley | Brzycki | Lombardi | Displayed mean | Formula spread |
|---|---|---|---|---|---|
| 100 × 5 | 116.67 | 112.50 | 117.46 | 115.54 | 4.96 |
| 100 × 10 | 133.33 | 133.33 | 125.89 | 130.85 | 7.44 |
Use case: compare strength records without moving the goalposts
Suppose an earlier same-exercise record is 100 × 5 and a later record is 105 × 5, both using the same equipment, unit, range-of-motion standard and logging rule. Because only load changes, every equation and their mean rise by exactly 5%. The displayed mean moves from 115.54 to 121.32, a difference of 5.78 load units. That is a clean mathematical comparison of two estimates.
The interpretation becomes weaker when the repetition counts, exercise variation, bar path standard, assistance, machine, grip, tempo, or set endpoint change. For example, comparing a paused bench press with a touch-and-go set or a free-weight squat with a machine squat changes the observed task. A precise percentage cannot make incompatible records equivalent.
Store the equation outputs or at least the calculator version with the record. If a later app silently switches formulas, an apparent strength change may be a model change. For several completed sets, use the multi-set estimated 1RM calculator so between-set disagreement remains visible instead of choosing whichever set produces the largest number.
- Keep exercise, unit, equipment and technique standard constant.
- Record whether the set was intended to be maximal for that repetition count.
- Compare individual equations as well as the displayed mean.
- Use a separate relative-strength calculation when body-mass normalization is genuinely the question.
Common mistakes, uncertainty and the decision boundary
Common mistakes include entering the total load on one date and per-side plate mass on another, counting assisted or incomplete repetitions as equivalent, comparing different exercises, and reporting the displayed mean as a tested personal record. Another mistake is hiding equation choice: two calculators can receive the same set and disagree simply because they implement different models.
Direct 1RM testing has shown good-to-excellent test–retest reliability across much of the published literature when a defined protocol is used, but reliability does not make every test safe, appropriate, or identical for every person. It also does not make an indirect estimate equivalent to a completed lift. Familiarization, exercise complexity, technique and the testing protocol remain relevant to what the number means.[2]
Do not use an estimate as an automatic next attempt, safe maximum, return-to-play decision, or individualized programme. Pain, illness, dizziness, recent injury, unfamiliar technique, equipment limits, spotting requirements and medical considerations sit outside these equations. Select actual loads only within an appropriate training and safety context.
- Estimated does not mean tested.
- The formula range is not a confidence interval.
- A higher estimate does not identify why performance changed.
- The calculator records arithmetic; it does not prescribe a load.
Practical questions
Frequently asked questions
Which one-rep max equation is the most accurate?
No equation is universally best for every exercise, repetition range and athlete. Compare the named outputs, keep the set context, and evaluate any equation against consistent tested or observed records relevant to the same exercise.
Is the mean of three estimates my true one-rep max?
No. It is only the arithmetic mean of the three displayed model outputs. It is not a measured lift, confidence interval, safety limit, or guarantee that the true result lies nearby.
Should I use a high-repetition set to estimate 1RM?
The calculator supports 1–20 repetitions, but a supported input is not a claim of equal predictive quality. Longer extrapolation and individual exercise differences can increase uncertainty, so keep the repetition count and equation spread visible.
Can I use the estimate to choose my next maximum attempt?
Not automatically. Attempt selection depends on technique, readiness, experience, equipment, spotting, rules and safety considerations that the calculator cannot observe. The result is a record and comparison aid, not a load prescription.
Further reading
Authoritative sources
Use these primary and professional resources to check definitions, conventions, or requirements that may extend beyond this guide.