Guide
Type a weight and a rep count into any one-rep-max calculator and it spits out a single number. But behind that number is a choice most tools hide from you: which formula did it use? There are half a dozen well-known equations for estimating a one-rep max (1RM), and they don't always agree. Here's how the main ones work, where each is most accurate, and why we average several instead of trusting just one.
Every 1RM formula is a curve fitted to real lifting data. It takes a submaximal set — say, 185 lb for 5 reps — and projects what you could theoretically lift for a single all-out rep. The projection works because there's a fairly predictable relationship between the percentage of your max you're lifting and the number of reps you can grind out at that weight. The formulas just express that relationship in slightly different math.
The catch: that relationship holds up well at low reps and gets fuzzy at high reps. Everyone agrees a heavy triple is close to your max. A set of 20, though, is influenced as much by your conditioning and pain tolerance as by raw strength — which is why estimates from high-rep sets are far less reliable.
The Epley formula (1985) is probably the most widely used. It's simple, and it's among the most accurate for sets of 1–5 reps. At exactly one rep it correctly returns the weight itself. As reps climb it rises in a straight line, which is why it can read slightly high on very high-rep sets.
Brzycki (1993) is the other everyday standard, and it maps neatly onto the classic "percentage chart" many lifters grew up with (e.g., a 5-rep max ≈ 87% of 1RM). It tends to read slightly lower than Epley at moderate reps, so the two make a nice bracket around a realistic answer.
Lombardi uses an exponent rather than a straight line. It stays closer to the others at low reps but climbs faster at high reps, so it can over-estimate from a set of 12+ — useful to know if a tool is quietly using it.
These three round out the set. O'Conner is a light linear formula; Wathan and Mayhew are exponential fits derived from bench-press research. None is dramatically better or worse than the others — they mostly differ in how aggressively they scale with reps. Wathan and Mayhew in particular were validated against bench-press data, so they tend to behave best on upper-body pressing and can read slightly differently on lower-body lifts where more total muscle is involved.
It helps to picture two families. Linear formulas (Epley, O'Conner) treat every extra rep as subtracting a fixed slice from the percentage of your max you're lifting, so their estimates climb in a straight line. Exponential formulas (Lombardi, Wathan, Mayhew) bend the curve, letting the rate of change taper or accelerate. At one to five reps the two families sit almost on top of each other — which is why low-rep estimates from different tools rarely disagree by much. It's only past roughly eight reps that the shapes visibly separate, and that separation is the single biggest source of "why did two calculators give me different numbers?" confusion.
The reason any of this works is a well-documented pattern strength coaches call the load–velocity or load–rep relationship: the heavier the load relative to your maximum, the fewer clean repetitions you can complete. Bodies such as the National Strength and Conditioning Association (NSCA) publish repetition-to-percentage tables built on exactly this idea, and every 1RM formula is a compact way of expressing the same curve. Because the relationship is population-averaged, your personal curve can be a little flatter or steeper — endurance-adapted lifters often grind out more reps at a given percentage than the tables predict, while pure strength specialists sometimes do fewer. That individual drift is a second reason to lean on low-rep sets, where the curve is steepest and personal variation matters least.
Take 225 lb for 5 reps. Here's what each returns:
| Formula | Estimated 1RM |
|---|---|
| Epley | 262 lb |
| Brzycki | 253 lb |
| Lombardi | 253 lb |
| Average of six | ≈ 257 lb |
At 5 reps the spread is only about 9 lb — small enough that it doesn't matter much. Push to 225 for 10 reps and the spread widens to 20+ lb, because the formulas disagree more the further they have to extrapolate. That's the real lesson: estimate from the lowest-rep set you reasonably can. A heavy set of 3 gives a far tighter estimate than a set of 12.
Three things quietly wreck an otherwise good estimate. The first is counting reps you didn't really earn — a rep with degraded form, a bounced bench, or a partial squat isn't the same rep the formulas assume, so padding your count inflates the result. The second is estimating from too many reps: as covered above, the formulas extrapolate further and disagree more the higher the rep count, so a set of 15 gives a much shakier number than a set of 4. The third is ignoring how you felt — the same weight for the same reps means something very different on a well-rested day than at the end of a hard week. The fix for all three is the same: estimate from a clean, honest, low-rep set taken on a day when you were reasonably fresh, and treat the output as a range rather than a precise figure. If two formulas disagree by 15 pounds, the truth is almost certainly somewhere between them, not at either extreme.
Because each formula drifts in a different direction as reps rise, no single one is "correct" across the whole range. Averaging several cancels out their individual biases and gives a more stable number than cherry-picking a favorite. That's exactly what our 1RM calculator does — it runs your input through six formulas and reports the average, while still letting you see the range.
Say you bench 185 lb for a hard set of 6 — the sixth rep was a real grind, meaning you had maybe one clean rep left in the tank. Run it through the formulas:
The six-formula average lands around 219 lb, with a spread of only about 7 lb. Now suppose you want your next top set to sit at 80% of your max: 0.80 × 219 ≈ 175 lb. You'd program 175 lb, expect it to feel like a solid but sub-maximal working weight, and never have to attempt a true single to get there. If, four weeks later, that same 185 moves for 8 clean reps, re-run the numbers — your estimated max has climbed to roughly 232 lb, and your 80% working weight should climb with it to about 186 lb. That loop — estimate, program from a percentage, re-estimate — is the entire practical point of these formulas.
For low-rep sets (1–5), the popular formulas agree within a few percent, so none is clearly "best." Epley and Brzycki are the most widely validated everyday choices. The more reliable move is to average several formulas rather than trust any single one, and to estimate from the fewest reps you reasonably can.
Aim for a heavy set of 3–5 reps taken close to failure. Fewer reps means less extrapolation and a tighter estimate. Sets above about 10 reps are heavily influenced by conditioning and pain tolerance, so estimates from them can be off by 10% or more.
Yes — that is exactly what it's for. Setting your training percentages from an estimate lets you avoid the injury risk of frequent true-max attempts. It is not a number you must go and lift; treat it as a planning figure and always warm up properly. This article is educational only and not a substitute for coaching or medical advice.
→ Estimate your 1RM with all six formulas