RPE vs RIR: what they mean, the conversion chart and how to use them

On the scale lifters use today, RPE and RIR say the same thing with a different number. RIR (reps in reserve) is how many more reps you could have done when you stopped. RPE (rating of perceived exertion) is that same set scored from 1 to 10. The conversion is the one Zourdos et al., 2016 used: RPE 10 is 0 RIR, RPE 9 is 1 RIR, and so on down.

The older RPE was different: a general score given after an exercise or a whole session. The reps-based version scores each set, which lets you adjust the weight from one set to the next (Helms et al., 2016).

RPE to RIR conversion chart

The rule fits on one line: RPE = 10 − RIR.

The whole-number rows come straight from the rule. The half points are a working convention for when you are torn between two numbers.

Add the reps you did to the reps you had left and you get your estimated rep maximum at that weight. The Epley formula turns that into a percentage of your 1RM. The sums in this table are ours, and they are an estimate rather than a measurement.

How to use it in a set

1. Set the target before you start: say, 3 sets of 8 at RPE 8, which means finishing with 2 reps in reserve.
2. Do the first set with a cautious weight. At the eighth rep, work out how many more you could do with the same form.
3. Write down weight, reps and RIR. Without them you have nothing to compare next week.
4. If you landed well away from the target, change the weight for the next set.

A worked example

You bench press 60 kg. You finish 8 reps and reckon you had 4 left: RPE 6, some way off the RPE 8 you wanted. That makes 8 + 4 = 12 as your estimated rep maximum, so your estimated 1RM is 60 × (1 + 12/30) = 84 kg. To do 8 with 2 in reserve you need a weight you could lift 10 times: 84 / (1 + 10/30) = 63 kg. On the bar that means 62.5 kg, and the plate calculator shows how to load it.

As a shortcut, on this formula each RPE point is worth moving the weight by 2 to 3%. If you would rather skip the sums, the one-rep max calculator gives you a starting point.

How well people judge it

Fairly well in short sets near failure, poorly far from it. The widest review we have read, Halperin et al., 2022 (12 studies, 414 people), found that people underestimated by 0.95 reps on average (95% CI 0.17 to 1.73), with a lot of variation between studies. Put simply, you usually have a bit more left than you think.

According to Halperin et al., 2022, people judge it a little better closer to failure, with heavier loads and in later sets, and training experience did not seem to matter. In a small study, nine trained men did not reduce their error significantly over six weeks (Remmert et al., 2023). And your 10 may not be someone else's: in Zourdos et al., 2016 (29 people, squat), experienced lifters scored their 1RM higher than novices did.

In practice, trust the number more in sets of 12 reps or fewer and at RPE 8 or above. «I had 5 left» in a set of 20 is a guess. Taking the odd set to failure to check yourself is a coaching habit, not something shown to work; if you do it, use a machine or have someone spot you.

What getting close to failure does

Refalo et al., 2023 pooled 15 studies measuring muscle growth in healthy adults of any age and training background.

The authors call the first advantage trivial and conclude there is no evidence that momentary failure beats stopping short for muscle growth. They point to a possible non-linear relationship: being closer to failure does not always mean more muscle.

The abstract does not say how many reps from failure is too far, and it looks at muscle growth, not strength. Leaving 1 or 2 reps on most sets is the usual advice: it fits these data, but it is a convention, not a figure from this analysis.

In GoForIt

GoForIt is not out yet: it is listed as «Coming soon». It estimates 1RM with this same Epley formula and shows working-set volume and personal records.