Where 1RM calculators quietly disagree
Load 225 lb for 5 reps and Onemax returns three different maxes off that one set: Epley reads 263 lb, Brzycki reads 253 lb, Lombardi reads 264 lb. None of them measured anything. Each is a separate published equation guessing at a single you never actually performed, and the 11 lb gap between the highest and lowest read is the part most calculators keep off screen. This page is about that gap: where it comes from, what the rest of the category does with it, and what Onemax does instead.
How a 1RM calculator usually decides your max
Most 1RM tools take the same two inputs you just gave, weight lifted and clean reps, and feed them into one closed-form equation that an exercise physiologist fit to a population of lifters decades ago. The tool picks a formula, hardcodes it, and prints the result like a scale reading. Strength Level does this well for its crowd and carries real trust with serious lifters, but its 1RM page runs a single formula and steers you toward its percentile-ranking product. ExRx.net has been a reference in the field for years and lists the equations honestly, yet it gives you no unit toggle and no way to switch the formula driving the number. Calculator.net pulls enormous traffic on generic authority and does offer three formulas plus a unit switch, then buries the formula choice and treats “save” as a browser bookmark. The pattern underneath all of them is the same: each one hardcodes a single equation, prints it like a settled answer, and never lets on that a different equation would have handed you a different number.
Three formulas, three different answers
Onemax ships three equations and lets you flip between them on one screen. With w as the weight lifted and r as clean reps:
1RM = w · (1 + r/30)1RM = w · 36/(37 − r)1RM = w · r^0.10They are built on different shapes. Epley, published by Boyd Epley in 1985, scales in a straight line, adding a flat one-thirtieth of the working weight per rep, so it keeps climbing without ever leveling off. Brzycki, from Matt Brzycki in 1993, divides by a denominator, 37 − r, that shrinks as reps rise; near r = 37 that denominator approaches zero, so Onemax guards it to keep the estimate finite instead of letting it run to infinity or flip negative. Lombardi, named for V. Patteson Lombardi's 1989 equation, raises reps to the 0.10 power, a curve that flattens fast and stays the most conservative of the three once reps hit double digits. Run the same 225 lb set through all three and the disagreement stops being academic:
| Reps | Epley | Brzycki | Lombardi | Spread |
|---|---|---|---|---|
| 1 | 233 | 225 | 225 | 8 |
| 3 | 248 | 238 | 251 | 13 |
| 5 | 263 | 253 | 264 | 11 |
| 8 | 285 | 279 | 277 | 8 |
| 10 | 300 | 300 | 283 | 17 |
| 12 | 315 | 324 | 288 | 36 |
Two rows are worth staring at. At a single rep, Epley returns 233 from a 225 lb input, not 225, because w · (1 + 1/30) is 1.033 · w; Brzycki and Lombardi both hand back the exact weight you lifted. At 10 reps Epley and Brzycki land on the identical 300, which is a coincidence and not agreement, since both reduce to 1.3333 · w at that one rep count and split apart again on either side of it.
What Onemax does with the disagreement
We default to Epley because it is the formula most lifters recognize, even though it reads high from the very first rep. That default is paired with a caveat that fires by rep count at the moment you calculate: clean through 1–5 reps, a caution flag around 6–10, and a low-confidence read past roughly 10 reps. When we ran competitor tools through sets of 12 and 15 reps while building this one, most printed an Epley-based number and said nothing about how far it had drifted, so we wired the caveat to call out Epley's overestimation by name when Epley is the active formula at a high rep count, rather than hedging in the abstract.
That confidence read is the one thing no competitor in this category surfaces at the point of calculation.
Onemax also does the plain parts the incumbents skip. All three formulas compute on one no-scroll screen, so you can watch the number move as you switch equations instead of trusting a single black box. An lb/kg toggle sits next to the inputs, which ExRx.net and Symmetric Strength both lack, and the %1RM table redraws from 50–100% in 5% steps every time the estimate changes. Gravitus carries more formulas than we do and a deeper feature set, but it lives behind an app-install funnel; Symmetric Strength does solid strength-standards work with no unit switch at all. The trade we made is a narrow feature count in exchange for one screen that shows its own uncertainty.
The trade-off you inherit above ten reps
Read down the spread column and the deal turns concrete. At 1 rep the three formulas sit 8 lb apart; at 5 reps, 11 lb; by 12 reps they fan out to 36 lb, with Brzycki reading 324 while Lombardi reads 288. The disagreement is smallest exactly where the formulas are trustworthy and largest exactly where they are not, because all three were fit mostly to low-rep data and are extrapolating once you push past a heavy set of about 5. Brzycki's shrinking 37 − r denominator is what drives it above the others at high reps, the reverse of its slightly-lower reputation in the low-rep range. The inheritance is easy to state and easy to forget: a 1RM projected from a set of 3 is a number you can program against, and a 1RM projected from a set of 15 is a rough direction. Onemax cannot make the high-rep math tighter than the equations allow, so it tells you on screen when the number has loosened. Re-test with a heavier set in the 1–5 rep range whenever you want that spread to close back down.
Common questions
Which one-rep max formula is the most accurate?
No single formula wins everywhere. Epley, Brzycki, and Lombardi were each fit to different lifter data, so they agree closely at low reps and drift apart as reps climb. For sets of one to five reps the three land within a few pounds of each other. Past ten reps Brzycki tends to read lower and Epley reads higher, so treat the number as a range rather than one fixed figure and lean on the formula that has tracked your own training sets best.
Why does the estimate change when I switch formulas?
Each formula is a different curve fit to a different data set, so they weight rep count differently. Epley adds a flat one-thirtieth bump per rep, Brzycki divides by a shrinking denominator, and Lombardi raises reps to a small power. At two or three reps those shapes nearly overlap. As reps rise the curves separate, which is why a ten-rep set can produce three noticeably different estimates from the same weight.
Can I trust a one-rep max estimate from a high-rep set?
The further your set sits from a true single, the looser the estimate becomes. All three formulas were fit mostly to low-rep data, so a fifteen or twenty rep set stretches them well past their comfortable range. Epley in particular is known to overestimate at high reps, often reading several percent above a real max. Use high-rep estimates as a rough ballpark for programming, and re-check with a heavier, lower-rep set when you want a tighter number.
What rep range gives the most reliable one-rep max estimate?
Sets of two to five reps give the tightest estimates. That range is close enough to a true single that the regression curves still track well, but far enough that you get a clean, repeatable set to feed the formula. A hard triple is the sweet spot for most lifters: low enough that Epley, Brzycki, and Lombardi agree, high enough to log without a maximal single every session.
How do I use the percentage-of-one-rep-max table for programming?
The %1RM table turns your estimate into working weights. Percentage-based programs prescribe sets at a share of your max, so read down to the percentage your program calls for and use that weight for the day. A 5/3/1 top set might sit at 85 to 95 percent, while a lighter volume day might call for 65 to 75 percent. Because the table recalculates from your live estimate, changing the input weight or formula updates every working number at once.
Does this one-rep max calculator save or send my numbers?
No. The calculator runs entirely in your browser and keeps nothing. There is no login, no account, and no server call, so your weight, reps, and formula choice never leave the page. Refreshing clears everything. It is built as a quick reference you can open, use, and close without leaving anything behind.
