// Strength Calculators

RPE Calculator

Give it your max, the reps you're planning, and either an RPE or the reps you mean to leave behind. It returns the load.

Your details

Your result

Load for that set
82.2kg
% of 1RM
82.2%
Reps in reserve
2
Treated as a set of
7 to failure
  • The percentage is not read from an RPE table. It is the same rep-max model the 1RM calculators here use, inverted on your reps plus your reps in reserve — so the tool is saying that a set of these reps at this effort resembles a set of that many taken to failure.
  • RPE is a judgement, not a measurement, and treating its output as precise is the mistake to avoid. Estimates of reps in reserve tend to be more reliable on heavier loads and closer to failure, and get looser on light loads and long sets.
  • If the load feels wrong in the gym, the load is right and the estimate is wrong. Adjust and re-derive from the set that actually happened.

Estimates are for general guidance only and are not medical advice. For a personalised plan, book a free guided tour at RPM Gym.

What your result means

The headline is a weight: what to put on the bar for the reps and the effort you asked for. Under it sit the percentage of your 1RM that load represents, the same effort written on the other scale, and the set the tool treated yours as.

That last line is the method, and it deserves a read rather than a glance. Five reps at RPE 8 is handled as a set of seven taken to failure, because RPE 8 means two reps left. The percentage then comes from running the rep-max model backwards on those seven reps — the same mean of Epley and Brzycki the 1RM calculators here run forwards. No grid sits behind this page.

Using that same model in both directions is the point, not a detail. A max estimated on one page and spent on another has to survive the trip: 100 kg for five estimates a 114.6 kg max, and asking this page for a five-rep load at RPE 10 hands back 100 kg. It used to hand back 98.2, because this page inverted only half the model the estimate was built from.

The RPE percentage charts in circulation would have been easier. They are also untraceable: they pass from programme to programme without a published derivation, and a table of unattributed percentages reads as more authoritative than the equations it displaced. Epley and Brzycki are named, dated equations with stated forms and known weaknesses, and those weaknesses can be printed on the page. That trade is worth making even though the output looks less tidy.

One case is fixed rather than modelled. A single at RPE 10 is 100% of your 1RM by definition, and that is what comes back — anything less would be the tool telling you your max is not your max.

How it is calculated

Epley and Brzycki averaged, inverted on reps plus reps in reserve

Reps in reserve = 10 − RPE (RPE 10 = 0 RIR, RPE 9 = 1, RPE 8 = 2, and so on) Effective reps = reps you plan + reps in reserve Epley load = 1RM ÷ (1 + effective reps ÷ 30) Brzycki load = 1RM × (37 − effective reps) ÷ 36 Load = the mean of those two, fixed at the 1RM itself for one effective rep % of 1RM = load ÷ 1RM

What it assumes

  • The 1RM you type is roughly right. Everything after it is a proportion of that number, so a max entered 4% optimistic returns loads 4% too heavy at every rep count on the page.
  • RPE and reps in reserve are one statement in two vocabularies. RPE 10 is nothing left, RPE 9 is one rep left, RPE 8 is about two — the anchoring that the resistance-training version of the scale was published with, and the reason this tool can take either input and hand back the other.
  • The rating counts reps, not breathing. What is being estimated is how many more you could have completed at the same technique, which is a different question from how unpleasant the set was.
  • Both equations were fitted to sets taken to failure. Folding reps in reserve into the rep count is what lets them read a set that stopped short, and it is the same move the 1RM pages make running the other way.
  • The curve is a group average. Rep-strength profiles differ enough that two lifters with the same max will not get the same number of reps at 80%, and nothing on this page can see which of them you are.

Units and rounding

  • 1RM in kilograms, 1 to 600, in half-kilo steps. The load returns in kilograms to one decimal place.
  • Reps you plan: whole numbers, 1 to 20. RPE runs 6 to 10 in whole points; reps in reserve runs 0 to 4.
  • At one effective rep the percentage is pinned to 100%, so a single at RPE 10 returns your max unchanged.
  • Pounds work without modification, since the output is a proportion of whatever unit you entered — but the answer is then in pounds, and one mixed entry inside a programme puts a working weight out by a factor of 2.2.

Worked example

A 140 kg squat max, five reps planned, RPE 8.

  1. RPE 8 → 2 reps in reserve, so the set is treated as 5 + 2 = 7 to failure
  2. Epley: 140 ÷ (1 + 7 ÷ 30) = 140 ÷ 1.2333 = 113.5 kg
  3. Brzycki: 140 × (37 − 7) ÷ 36 = 140 × 0.8333 = 116.7 kg
  4. Load = the mean of the two = 115.1 kg, which is 82.2% of the max
  5. The same five reps at RPE 9: 5 + 1 = 6 → 116.7 and 120.6 → 118.6 kg, or 84.7%

115.1 kg for five at RPE 8; 118.6 kg for five at RPE 9. A whole point of RPE is worth about 3.5 kg here, or 2.5% of the max — roughly the size of the error a pooled analysis found in lifters' own predictions of the reps they had left. Load the number, then let the set correct it.

Limitations

  • RPE is a judgement, not a measurement. Nothing is being sensed: you estimate how many reps were still there, and the tool turns that estimate into a weight quoted to a tenth of a kilogram. The decimal belongs to the arithmetic, not to the guess it was built on.
  • The estimate leans one way. A review pooling twelve studies of people predicting their own reps to failure found they undershoot by about one rep on average, which means a set you file as RPE 8 has a decent chance of being a 7.
  • The same review found predictions tighten closer to failure, on heavier loads and on shorter sets — under about twelve reps. Long sets at light loads are where the judgement is worst and where both equations are also furthest from the data they were fitted to. Two sources of error pointing the same direction.
  • Averaging the two equations is a choice, not a finding. Epley and Brzycki were each fitted and published on their own; nobody validated their mean as a third model. It is used here because the two disagree in opposite directions as the reps climb, so the average is the steadier of the options — and because using it in both directions is what makes a max estimated on the 1RM page spend correctly on this one. That is a consistency argument, not evidence that the blend is more accurate than either equation alone.
  • The percentage does not travel between exercises. Rep tolerance is exercise-specific, and the tool never asks what you are lifting.
  • It cannot see the day. Sleep, food, and a heavy session 48 hours ago all move what RPE 8 weighs, and none of them are inputs. If a lift hurts, the effort scale is the wrong instrument entirely — pain is not a rating.

What to do next

  • Load the number, then rate the set afterwards. If the reps you genuinely had left do not match what you asked for, the 1RM is the input to fix — the arithmetic between the two is not in question.
  • Enter the scale you actually think in. Someone writing 3 × 5 @ 8 and someone writing 3 × 5 with two left are asking this tool the same question, and it answers in either language.
  • Recalculate when the max moves, not after a session that felt good. The load here is a fixed proportion of the 1RM you feed it, so nothing changes until that number does.
  • Test your own bias every few weeks. Finish a set you would have stopped, count what you got past your predicted stopping point, and keep the difference. Doing that on a machine costs less than discovering it under a barbell.
  • On sets longer than about ten reps, treat the output as a bracket rather than a target. Round to what the bar can actually be loaded to and spend the attention on the reps instead.

Sources

Related calculators
View all calculators →
// Frequently Asked

Questions, answered.

What does RPE mean in weight training?

How close a set came to failure, counted in reps rather than in feeling. RPE 10 means the next rep was not there. RPE 8 means about two more were. The resistance-training version of the scale is anchored to reps in reserve for exactly that reason: hard is not a number, but two left is.

What is the difference between RPE and RIR?

The same information counted from opposite ends. RIR counts up from zero, RPE counts down from ten, and 10 minus your RPE is your RIR. That is why there is no separate reps-in-reserve tool to build — enter either one here and the other is printed next to your load.

Why is there no RPE percentage chart on this page?

Because nobody can show where the numbers in one came from. Those charts are practitioner tables that got copied between programmes, and the version you find depends on who you asked. An equation with a citation behind it can be checked, argued with and corrected. A chart with no author cannot, and it would sit on this page looking more precise than the thing it replaced.

Is RPE 8 the same weight every week?

No, and the tool has no way to warn you. It returns the load your max implies for that effort, which is a fixed number until the max changes. What RPE 8 actually weighs on a given day moves with sleep, food, stress and what you trained two days earlier. Set the bar from the output and let the first working set adjust it.

Can I enter half points like RPE 8.5?

Whole points only, because the reps-in-reserve anchor the tool converts through is a whole rep. Half a point sits between one rep left and two, and on five reps it comes to roughly 1.8 kg on a 140 kg max — under the smallest jump most gyms can load. Write halves in your log if they help you think; the arithmetic will not honour them.

Can I get better at judging reps in reserve?

The validation work points that way: experienced squatters rated a maximal single more accurately than novices, and the link between their ratings and bar speed was tighter. The practical version is to occasionally finish a set you would have called done and count what you were wrong by. Most people find the gap narrows once they have measured it a few times.

Should I programme with RPE or with percentages?

A percentage fixes the load and lets effort float; an RPE fixes the effort and lets the load float. This page is the bridge, printing one beside the other so a programme written in either can be run in either. As a rule of thumb, percentages are easier to plan in advance and RPE is easier to execute on a day that did not go to plan.

The calculated weight felt far easier than the RPE I asked for. What went wrong?

Two candidates, and they are distinguishable. If it feels light at every rep count you try, the 1RM you entered is low — re-estimate it from a recent set. If it feels light at five and brutal at twelve, that is your own rep-strength profile rather than a fault in the equation, and the fix is a personal offset at the high-rep end rather than more decimal places.

RPE 8 on Tuesday is not RPE 8 on Saturday.

Sleep, the week behind you and the way the first warm-up moves all shift where that effort lands, and adjusting the bar mid-session is a judgement worth borrowing before you own it. Book a free guided tour at RPM Gym in Al Manhal.

Book Free Tour