Why does the formula change when I switch lifts?

Because different lifts have different fatigue curves. Bench press shows clean linear drop-off, so Mayhew (with Epley as a sanity check) tracks best. Squat's lower-body endurance lets lifters grind out reps standard formulas can't predict, Wathan's slower-decaying exponential handles it better. Deadlift's static start and neural fatigue make every formula underpredict. OHP and isolation lifts fail abruptly because the muscle mass is small, so Brzycki's steeper decline matches the reality. Picking one formula across all lifts is the lazy default; per-lift is what the sports-science literature actually supports.

Why are the rep caps different per lift?

Past the per-lift cap, the rep-to-1RM relationship breaks down. Bench and squat hold up to about 8 reps because larger muscle masses sustain submaximal work. Deadlift falls apart past 5 because grip and neural fatigue dominate. Overhead press caps at 5 because the smaller delt/triceps fail abruptly. Isolation lifts cap at 5 because metabolic burn (lactic acid) takes over long before mechanical failure, so a 12-rep curl set tells you almost nothing about your 1-rep curl.

Why is deadlift always underpredicted?

Two reasons. First, the deadlift starts completely static, there's no eccentric (lowering) phase to load elastic energy, so each rep is a near-cold start. Second, grip and CNS fatigue stack faster on deadlift than any other lift. LeSuer et al. 1997 found every common formula significantly underpredicted real deadlift maxes. The honest move is to treat the calculator number as a conservative floor and expect your real 1RM to land 5–10% higher.

Why is leg press handled differently?

Machine stability removes core and stabilizer demands, lifters can grind out two to three times as many reps at high percentages of 1RM on a leg press as on a free-weight squat. Standard formulas badly overpredict if you feed them a 12-rep leg-press set. The Adjusted Epley used here (multiplier 0.0242 instead of 0.0333) is a published correction for that, validated against Nuzzo et al. 2023 meta-regressions. Stay within 8 reps and the number will be honest.

Should I bother with 1RM math on bicep curls?

Honestly? No. Isolation lifts are hypertrophy work. You don't program curls or pushdowns off a 1RM, you program them off rep targets, RPE, and weekly volume. The Isolation option exists in this tool because the research has a defensible (Brzycki, ≤5 reps) answer for the rare case where you want one, not because it's a number most people should care about.

What's the safest way to actually test my real 1RM?

Don't, unless you're prepared. Real max testing wants 1–2 weeks of peaking, a competent spotter or a power rack with pins set at fail-height, and a warm-up ramp that builds in 5–10% jumps. Most lifters get more useful information from rep-PRs and AMRAPs than from going for a true single.

How do I use the training-percentage table?

Pick the row that matches the goal of your block. Building strength: live in the 80–90% zone for sets of 3–6. Building muscle: 70–80% for sets of 6–12. Conditioning or warm-ups: 60–70% for higher reps. The weights shown are rounded to plate-friendly increments (5 lb or 2.5 kg) so you don't end up loading 187.5 lb on a bar.

All tools

Lift-aware 1RM calculator

The formula should know which lift you did.

Bench, squat, deadlift, OHP, leg press, and isolation lifts each have a different fatigue curve. Most apps run one formula across all of them. This one picks the research-recommended formula and rep cap per movement, and tells you the bias.

Your set

Lift

Each lift uses the formula sports-science literature shows tracks best for that movement’s fatigue curve. See the caveats per lift below.

Weight lifted

The actual weight on the bar or the machine’s loaded total.

Reps performed

Real set to honest failure or RPE 9. Max for bench press: 8 reps (research-backed accuracy ceiling for this movement).

Don’t actually max out on a calculator result. Heavy singles are a skill, attempt them under supervision (or a power rack with pins) after specific peaking work.

Bench Press

Your estimated bench press 1RM

Mayhew · preferred for bench press

(100 × w) ÷ (52.2 + 41.9 × e^(−0.055 × r))

268lb

(100 × w) ÷ (52.2 + 41.9 × e^(−0.055 × r))

Epley · sanity check

w × (1 + r/30)

263lb

Behavioral quirk · bench press

Linear, predictable fatigue. The Mayhew formula tracks the cleanest within 8 reps; Epley is shown alongside as a sanity check.

Training percentages

From the Mayhew estimate, rounded to plate-friendly weights (nearest 5 lb).

% of 1RM	Weight	Typical reps	Used for
95%	255 lb	1–2	Singles, peak strength
90%	240 lb	3–4	Heavy triples
85%	230 lb	5–6	Strength work
80%	215 lb	8	Heavy hypertrophy
75%	200 lb	10	Classic hypertrophy
70%	185 lb	12	Volume / endurance
65%	175 lb	15	Conditioning / warm-ups

Re-test off a real set every 4–8 weeks rather than chasing the calculator. The primary formula is the research-recommended single best estimate for this lift; the sanity check (when shown) gives you a feel for the spread.

The math

Different lifts, different formulas.

Why a submaximal set works. There’s a predictable relationship between the heaviest weight you can lift for one rep (1RM) and lighter weights you can lift for more reps. If you bench 200 for 5 clean reps to failure, that set encodes your 1RM. The catch: how reps drop off as you approach maximum isn’t the same across lifts. Lower-body endurance is higher than upper-body. Deadlifts fatigue the nervous system. Machines remove stability demand. One formula across everything is the lazy default; the literature supports better.

The five formulas this tool uses.

Epley:           1RM = w × (1 + r/30)
Brzycki:         1RM = w ÷ (1.0278 − 0.0278 × r)
Mayhew:          1RM = (100 × w) ÷ (52.2 + 41.9 × e^(−0.055 × r))
Wathan:          1RM = (100 × w) ÷ (48.8 + 53.8 × e^(−0.075 × r))
Adjusted Epley:  1RM = w × (1 + 0.0242 × r)

Epley and Brzycki are linear (each rep takes a fixed fraction off your max). Mayhew and Wathan are exponential (the decline is steep at first, then flattens). Adjusted Epley is Epley with a smaller multiplier, calibrated for machine work where reps inflate compared to free weights.

Which formula for which lift.

Lift	Primary	Sanity check	Rep cap
Bench Press	Mayhew	Epley	8
Squat	Wathan	Epley	8
Deadlift	Epley	—	5
Overhead Press	Brzycki	—	5
Leg Press	Adjusted Epley	—	8
Isolation (curl, pushdown)	Brzycki	—	5

What kills accuracy.Sets that weren’t actually to failure (or honest RPE 9). Form breakdown on the last rep. Holding back because of fear or a missing spot. Going past the per-lift cap. Using the number from a single set instead of trending it across multiple weeks.

How to actually use it. Take a top set every 4–8 weeks. Plug it in. Program the next block off the primary estimate and the working-weight table below the result. Re-test off a real set when the block ends, don’t chase the calculator number.

Worked example

Two lifts, two answers.

Bench press · 225 lb × 5. Primary (Mayhew): (100 × 225) / (52.2 + 41.9 × e^−0.275) ≈ 22,500 / 84.0 ≈ 268 lb. Sanity check (Epley): 225 × (1 + 5/30) = 263 lb. Tight cluster, the 5-lb gap is the window your real 1RM lives in. Program off 268.

Deadlift · 315 lb × 5. Primary (Epley): 315 × (1 + 5/30) = 368 lb. No sanity check on deadlift, every common formula underpredicts because the static start and CNS fatigue drop rep capacity faster than the math expects. Treat 368 as a conservative floor; your real deadlift 1RM is likely in the 380–405 range.

From those numbers.80% of 268 lb bench ≈ 215 lb for sets of 8 (heavy hypertrophy lane). 85% of 368 lb deadlift ≈ 315 lb for sets of 5 (strength work, conveniently your test set). Don’t chase the calculator; chase the trend across blocks.

FAQ

Honest answers.

Why does the formula change when I switch lifts?: Because different lifts have different fatigue curves. Bench press shows clean linear drop-off, so Mayhew (with Epley as a sanity check) tracks best. Squat's lower-body endurance lets lifters grind out reps standard formulas can't predict, Wathan's slower-decaying exponential handles it better. Deadlift's static start and neural fatigue make every formula underpredict. OHP and isolation lifts fail abruptly because the muscle mass is small, so Brzycki's steeper decline matches the reality. Picking one formula across all lifts is the lazy default; per-lift is what the sports-science literature actually supports.
Why are the rep caps different per lift?: Past the per-lift cap, the rep-to-1RM relationship breaks down. Bench and squat hold up to about 8 reps because larger muscle masses sustain submaximal work. Deadlift falls apart past 5 because grip and neural fatigue dominate. Overhead press caps at 5 because the smaller delt/triceps fail abruptly. Isolation lifts cap at 5 because metabolic burn (lactic acid) takes over long before mechanical failure, so a 12-rep curl set tells you almost nothing about your 1-rep curl.
Why is deadlift always underpredicted?: Two reasons. First, the deadlift starts completely static, there's no eccentric (lowering) phase to load elastic energy, so each rep is a near-cold start. Second, grip and CNS fatigue stack faster on deadlift than any other lift. LeSuer et al. 1997 found every common formula significantly underpredicted real deadlift maxes. The honest move is to treat the calculator number as a conservative floor and expect your real 1RM to land 5–10% higher.
Why is leg press handled differently?: Machine stability removes core and stabilizer demands, lifters can grind out two to three times as many reps at high percentages of 1RM on a leg press as on a free-weight squat. Standard formulas badly overpredict if you feed them a 12-rep leg-press set. The Adjusted Epley used here (multiplier 0.0242 instead of 0.0333) is a published correction for that, validated against Nuzzo et al. 2023 meta-regressions. Stay within 8 reps and the number will be honest.
Should I bother with 1RM math on bicep curls?: Honestly? No. Isolation lifts are hypertrophy work. You don't program curls or pushdowns off a 1RM, you program them off rep targets, RPE, and weekly volume. The Isolation option exists in this tool because the research has a defensible (Brzycki, ≤5 reps) answer for the rare case where you want one, not because it's a number most people should care about.
What's the safest way to actually test my real 1RM?: Don't, unless you're prepared. Real max testing wants 1–2 weeks of peaking, a competent spotter or a power rack with pins set at fail-height, and a warm-up ramp that builds in 5–10% jumps. Most lifters get more useful information from rep-PRs and AMRAPs than from going for a true single.
How do I use the training-percentage table?: Pick the row that matches the goal of your block. Building strength: live in the 80–90% zone for sets of 3–6. Building muscle: 70–80% for sets of 6–12. Conditioning or warm-ups: 60–70% for higher reps. The weights shown are rounded to plate-friendly increments (5 lb or 2.5 kg) so you don't end up loading 187.5 lb on a bar.

Modern alternative

Velocity-based training, when you want real precision.

Rep-based formulas estimate. Velocity-based training (VBT) measures. An accelerometer or camera tracks bar speed (in meters per second) across your warm-up sets. The Load-Velocity Relationship is so reliable that a linear regression chart can pinpoint exactly the weight at which bar speed will drop to your Minimal Velocity Threshold, the speed a 1RM travels at.

Empirically: a bench press 1RM typically moves around 0.15 m/s; a back squat 1RM lands near 0.30 m/s. You can predict 1RM without ever attempting it. The downside: it needs hardware. Devices like the GymAware or open-source smartphone tools work, but the math here is free.

For most lifters, the per-lift estimates above are accurate enough to program a real block. VBT is the upgrade when ±5% isn’t close enough, peak comp prep, sport science settings, elite-level programming.

References

What this tool draws on.

Formula-per-lift mappings and the rep-cap thresholds reflect peer-reviewed sports science, not internet rules of thumb.

LeSuer 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. Foundational comparison of Brzycki, Epley, Lander, Mayhew, O’Conner, and Wathan formulas across the three barbell lifts. Source of the per-lift accuracy preferences this tool uses.
Nuzzo et al. (2023). Meta-regression on the repetition-to-failure relationship across resistance exercises. Showed why machine work (leg press) inflates rep counts compared to free-weight equivalents, motivating the Adjusted Epley multiplier.
Macarilla et al. (2022). Documented the breakdown of prediction-equation accuracy above 10 reps and the Minimal Velocity Thresholds referenced in the VBT callout above (~0.15 m/s bench, ~0.30 m/s back squat).
O’Connor (2015). Practical case for submaximal estimation over true 1RM testing in athlete programming, the injury-risk and fatigue argument this calculator is built on.
Marston et al. (2022). Load-Velocity Relationship modeling for individualized strength prediction; the paper underpinning the VBT section.

Want the program behind the percentages?

Knowing your estimated 1RM is the easy part. Building the right block around it, picking the right lifts, the right loads, the right peak, is the part that actually moves the bar. That’s coaching.

Talk about working together

A note on the numbers. These tools use established formulas (Mifflin-St Jeor, AMDR macros, healthy-range references) and population averages. Treat results as a starting point, not personalized medical advice. Your individual needs may differ based on body composition, training history, and medical context. Consult a qualified healthcare professional before making significant changes to your nutrition or training, especially if you have a pre-existing condition. See our full fitness and medical disclaimer.