Bench Press 1RM Guide: How to Estimate Your True Max
The bench press is where 1RM calculators start the most arguments — small technique differences move the number more than on any other lift. Use a 3-to-5 rep set, done the same way every time (same pause, same arch, same grip), and Epley gets you close: 225 lb × 5 estimates to 262.5 lb.
Should I enter paused or touch-and-go reps?

Enter whatever you actually did — but know they’re different numbers. A paused rep, with the bar motionless on the chest for a beat, typically runs about 95% of a touch-and-go max for most lifters. That’s coaching consensus rather than lab science, but it holds up remarkably well across gyms.
The calculator has no idea which style you used. Feed it touch-and-go reps and it predicts a touch-and-go max. If you compete in powerlifting, where every rep is paused on command, test paused. A paused test set predicts your platform number far better than bouncy gym reps ever will.
And never mix them. Five paused reps and five touch-and-go reps are two different inputs that deserve two different estimates. Pick the style that matches your goals and be boring about it.
Does bouncing the bar ruin the estimate?
Completely. When the bar rebounds off your ribcage, the stretch reflex donates reps your pressing muscles didn’t earn. The calculator reads those inflated reps as genuine strength and hands back a max that exists only on screen.
You’ll discover the truth the first time you try to pause that estimated weight — it won’t move. Touch the chest under control on every rep of a test set. The number you get will be smaller and infinitely more useful.
How much does my arch matter?
A moderate arch is legitimate lifting. It puts your shoulders in a safer position and it’s legal in every federation. But it shortens the bar path, and a big arch combined with high reps is the classic recipe for an estimate that looks heroic on screen and humiliating under the bar.
Here’s the honest framing: your arch is part of your technique, and the estimate belongs to that technique. A lifter with a competition arch who estimates off arched reps gets a number that predicts arched singles. That’s fine. What’s not fine is estimating off a huge arch and then attempting the number flat-backed, or vice versa. Same setup, every test, no exceptions.
Does grip width change the estimate?
Yes, though less than depth changes a squat. A wider grip shortens the range of motion and usually lets you handle more weight; a close grip turns the lift into a triceps-dominant movement with a longer path. Competitive lifters often bench 10% or more less close-grip than with competition width — observed range, not a formula.
Same rule as everything else on this page: the estimate belongs to the grip you tested. Changing grip width between test sets is changing lifts. If you train multiple grips, that’s fine — just estimate each one separately.
Why did the calculator say 330 when I can’t hit 295?
This is the most common bench estimation story in existence, and it always has the same plot. A lifter does 225 for 15 reps. Epley says 225 × (1 + 15 ÷ 30) = 337.5. Brzycki is even more optimistic at 368. The lifter loads 330, unracks it, and it doesn’t budge. Real max: 295.
What happened? Fifteen reps is an endurance set wearing a strength costume. Past about 8-10 reps, the formulas extrapolate your strength curve way beyond what the data supports — and on bench, where technique breakdown compounds rep by rep, the extrapolation fails hardest. The community verdict is blunt: above 5 reps, bench estimates get speculative; at 15, they’re basically fan fiction.
The fix is simple and unglamorous: retest with 3 to 5 reps, close to failure, with competition-legal technique. That 225 × 5 gives Epley 262.5 and Brzycki 253.1 — a range you can actually train off instead of a fantasy to be disappointed by.
Worked example: 225 lb × 5
Epley: 225 × (1 + 5 ÷ 30) = 225 × 1.1667 = 262.5 lb. Brzycki: 225 × 36 ÷ (37 − 5) = 225 × 1.125 = 253.1 lb.
Nine pounds apart on the same set. Program off 253 — the bottom of the range — and you’ll build strength on weights you can actually handle. Program off 263 and you’re flirting with missed reps. When two formulas disagree, the conservative one is usually the better coach.
How do I use my bench estimate in training?
Convert it to a training max first — roughly 90% of the estimate, the way 5/3/1 does it. A 260-lb estimated max becomes a 234-lb training max, and your working percentages come off that number. This one step prevents more stalls and injuries than any other programming decision, because it stops you from training at your absolute ceiling every session.
Retest the estimate whenever you hit a strong 3-to-5 rep set in normal training — it’s free information. Save true max attempts for the end of a training block. And if your estimate climbs while your paused single stays put, the gap is telling you something about your technique, not your strength.
Frequently Asked Questions
Can I estimate my bench max from dumbbell bench numbers?
Only roughly. Dumbbells demand more stabilization, so dumbbell numbers run lower — expect your barbell max to sit something like 10–20% above your dumbbell estimated max, though individual variation is large. Enter the combined weight of both dumbbells, not one, or the math is wrong from the first step.
Why is my touch-and-go max so much higher than my paused max?
The stretch reflex. Touch-and-go reps recycle elastic energy rep to rep; a pause kills it dead. The ~5% gap between the styles is normal and expected. Neither number is “the real one” — they’re two real numbers for two different lifts.
How often should I test my bench 1RM for real?
A few times a year, at the end of a training block or before a meet. Between tests, let the calculator do the work: every strong 3-to-5 rep set is a fresh estimate without the recovery cost of a true single.
Do I need a spotter for the test set?
For a 3-to-5 rep estimation set taken near failure: yes, always. “Near failure” on bench means the bar might not come back up, and that’s exactly the rep that makes the estimate accurate. No spotter, no safety arms, no close-to-failure set. Non-negotiable.
My estimate hasn’t moved in months. Is the calculator broken?
The calculator is fine — your training probably needs adjusting. A flat estimate over 8–12 weeks usually means the program needs more volume, better exercise selection, or a deload you skipped. The estimate is a mirror. Don’t blame the mirror.
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), 211–213.
Learn more





