Deadlift 1RM: How to Estimate Your Max Pull
Your deadlift is the hardest of the big three to estimate. Every major formula undershoots it by 9–14% (LeSuer et al., 1997), your grip quits before your back does, and touch-and-go reps inflate the number. The honest way to estimate it: pull 3–6 dead-stop reps, run the Epley formula, and expect your real max to sit above the result.
Why is the deadlift the hardest lift to estimate?

Because the research says so, plainly. LeSuer and colleagues tested seven prediction equations on 67 untrained college students across the bench press, squat, and deadlift, and published it in the Journal of Strength and Conditioning Research in 1997. Every formula correlated strongly with tested maxes — all r values above 0.95, which sounds great until you read the fine print.
The fine print: every single formula underestimated the deadlift by 9–14%, or 22.2 to 33.7 pounds in absolute terms. Not one of them overestimated it. Not one was even neutral. The deadlift is the only lift where the entire family of equations leans the same wrong direction.
Why? Nobody proved a single mechanism, but the honest answer is that the deadlift is the least “clean” lift to model. There’s no stretch reflex at the bottom like a bench press has. Your grip is a separate failure point from your back and legs. And most test sets are done touch-and-go while real maxes are pulled dead-stop — the formula is comparing two different movements and pretending they’re one.
Practical takeaway: whatever number your calculator gives you for deadlift, your true max is probably sitting a bit above it. I’d plan training weights off the estimate and keep the difference in your back pocket for test day.
Is your grip the real limiter?
Often, yes — and the formula has no idea. I’ve watched lifters pull 500 for five reps who couldn’t budge 540 for a single, not because their back wasn’t strong enough but because their hands gave out. The equation just sees “500 × 5” and spits out a number. It doesn’t know your calluses were tearing.
This is the single biggest reason deadlift estimates feel “off” compared to bench or squat estimates. On those lifts, the muscles doing the work are the muscles failing. On deadlift, there’s a decent chance the weak link is your hands, and the formula is quietly estimating your grip’s max instead of your pull’s max.
So be honest with yourself about what ended the set. If your back and legs had more in them but the bar was sliding out of your fingers, the estimate is measuring the wrong thing. Either retest with straps and note it, or accept the number for what it is — a grip-limited estimate — and don’t program your back work off it.
Should you pull touch-and-go or dead-stop?
Dead-stop. Every time, for estimation purposes.
Here’s the difference: touch-and-go lets you bounce the plates off the floor and ride the stretch reflex into the next rep. That’s not cheating in training — it’s a legitimate style — but it inflates your rep count relative to a true dead-stop pull, and the formula assumes each rep started from a dead stop. Ten touch-and-go reps and ten dead-stop reps are genuinely different sets, and only one of them predicts a competition-style max.
The standard I use: lower the bar under control, let the plates settle completely, reset your brace, and pull. Don’t let go of the bar between reps — just pause long enough that the bounce is gone. If your test set looked like that, your estimate means something. If it was a bounce-fest, knock your expectations down a notch.
Does sumo vs conventional change the number?
They’re different lifts with different leverages, so treat them that way. Sumo shortens the range of motion and shifts more work to the hips and quads; conventional loads the back harder and travels further. Most lifters pull more one way than the other, and the gap varies wildly person to person — limb length and hip structure decide it more than any formula can.
The rule is simple: never mix them in one estimate. A conventional test set predicts a conventional max. A sumo test set predicts a sumo max. If you pull both stances, track two separate numbers the way you’d track bench and overhead press separately. The calculator doesn’t know which stance you used, so you have to.
What rep range gives the best deadlift estimate?
Three to six reps. Tighter than the general guidance for other lifts, and for good reason.
Past about eight reps on deadlift, something changes: your cardiovascular system starts tapping out before your muscles do. A set of twelve on deadlift is partly a conditioning test, and the formula — which assumes muscular failure ended the set — reads it wrong. This is also where grip fatigue compounds rep by rep, quietly turning the set into a grip test wearing a deadlift costume.
The research backs the low end too. Reynolds et al. (2006) found the 5-rep max the most accurate predictor tested, with an R² of 0.993 on bench — and while that study wasn’t deadlift-specific, the principle holds: the closer your test set is to a single, the less the formula has to guess. Three to six heavy, clean, dead-stop reps is the sweet spot. Singles and doubles work too if you’re experienced enough to hit them safely.
Straps or no straps on test sets?
Match your test to your goal. If you compete in a federation that doesn’t allow straps, test without them — a strapped estimate will overpromise on meet day when your bare hands are the only thing on the bar. If you always train with straps and just want a number for programming, strap up and be consistent about it.
What you must not do is mix: a strapped test set feeding an estimate you’ll attempt unstrapped. Pick one setup, note which one it was, and retest the same way every time. Consistency beats purity here — the trend over months matters more than any single number.
Do these formulas work the same for women?
Not exactly, and it’s worth knowing why. The prediction equations were built overwhelmingly on male data, and women tend to grind out more reps at near-max percentages than the curves expect. A woman hitting 5 reps at 90% is less unusual than the formula thinks it is — which means the estimate can overshoot her true single.
The fix is the same as everywhere else on this page: keep test sets in the 3–5 rep range, where the formulas are tightest for everyone, and calibrate against a real tested max when you get the chance. One honest data point — “my formula said 275, I pulled 260” — is worth more than any population curve.
Worked example: 315 lb × 5 reps
Let’s run a real set through both of the most common formulas so you can see the spread.
| Formula | Math | Estimated 1RM |
|---|---|---|
| Epley (1985) | 315 × (1 + 5/30) | 367.5 lb |
| Brzycki (1993) | 315 × 36/(37 − 5) | 354.4 lb |
Thirteen pounds apart from the same set — that’s normal, not a bug. Epley runs a little hotter at higher reps; Brzycki is the conservative one. I’d treat 354–368 as the honest range, plan training off the lower end, and remember the LeSuer finding: your real deadlift max is likely sitting above both numbers, not between them.
How do you turn the estimate into training weights?
Don’t program off 100% of an estimate — that’s stacking assumptions on assumptions. The standard move, straight from Jim Wendler’s 5/3/1, is to take 90% of your estimated max as a training max and base all your percentages on that. With our example: 90% of 360 (splitting the Epley/Brzycki range) is 324 lb. That’s your anchor.
From there it’s simple percentages of the training max: 65–75% for volume work, 80–85% for heavy sets, 90%+ when you’re testing or peaking. The NSCA’s rule of thumb helps sanity-check: your 10-rep weight should sit around 75% of your max. If the numbers feel wildly off from what you actually lift, trust the barbell over the spreadsheet and recalibrate.
And retest every couple of months — not every week. A fresh 3–6 rep set, same stance, same straps-or-not setup, run through the same formula. The trend is the program.
Deadlift 1RM questions, answered
Why is my deadlift estimate always low?
Because the formulas are biased that way — LeSuer et al. (1997) found every tested equation underestimated the deadlift by 9–14%. Grip giving out before your back is the other usual suspect. Expect your real pull to beat the estimate.
Should I test sumo or conventional?
Whichever you actually train. They’re different lifts with different leverages, so a sumo test set only predicts a sumo max. Track them separately and never blend the numbers.
Do I count a touch-and-go set?
You can, but know what it is: touch-and-go inflates rep counts via the stretch reflex, so the estimate will run hot versus a dead-stop max. For the most honest number, retest with full dead-stop reps.
How often should I retest my deadlift max?
Every 8–12 weeks is plenty. Heavy deadlift testing is the most fatiguing of the big three, and weekly maxing just digs a recovery hole. Between tests, re-estimate from a strong 3–6 rep training set.
Will my gym estimate match a competition pull?
Close, with caveats. Competition pulls are dead-stop, judged, and usually unstrapped — if your test set matched those conditions, the estimate transfers well. Add the LeSuer underestimation on top and most lifters find their meet PR lands a little above their gym estimate.
Sources
- LeSuer, D.A. et al. (1997). “The accuracy of prediction equations for estimating 1-RM performance in the bench press, squat, and deadlift.” Journal of Strength and Conditioning Research, 11(4). doi:10.1519/R-17094.1
- Haff, G.G. & Triplett, N.T. (Eds.). Essentials of Strength Training and Conditioning (4th ed.). NSCA / Human Kinetics.





