What is RPE in weight training, and can you trust it?

What is RPE in weight training?

RPE, short for rating of perceived exertion, is your rating of how hard a set felt, from 1 to 10. In lifting, it’s often counted in the reps you had left before failure: 10, none left; 9, one; 8, two. This scale was tested on the squat in 29 people, 15 experienced lifters and 14 novices; its authors see it as a practical way to set the day’s weight (Zourdos et al., 2016).

Where does what you feel during a set come from?

Partly from the command your brain sends to your muscles. In 16 men lifting a light then a heavier weight with one-arm elbow flexions, with a fresh or a fatigued arm, perceived effort rose with the weight and with fatigue, together with a brain signal that reflects that command, recorded on the scalp; the stronger one was, the stronger the other (de Morree et al., 2012). As your muscle tires, it takes a stronger command to lift the same weight, and you feel it. That’s a measured link, not the whole story: fatigue in the muscle itself may play a part too (Guo et al., 2017).

And the bar slows down. In the 29 people tested on the squat, from 60% of their max up to their max, the slower the bar moved, the higher their rating (Zourdos et al., 2016). Same finding in 27 trained lifters on the front squat and the hex bar deadlift, in sets at 80% of max taken to failure (Odgers et al., 2021). These are measured links, not proof that one causes the other.

Can you trust your RPE?

Roughly, to within about one rep. One analysis pooled 12 studies and 414 people: each one predicted how many more reps they could do, then went to failure. On average, they guessed 0.95 reps fewer than they had left, with large differences between studies; training status didn’t seem to matter (Halperin et al., 2022), though a study of 141 people found a tendency to get more accurate with experience (Steele et al., 2017). The analysis authors leave open whether that error is acceptable.

The guess gets a little more accurate near failure, on heavy sets of 12 reps or fewer, and in later sets; beyond 12 reps, the error grows much faster with set length (Halperin et al., 2022). In the 27 trained lifters at 80% of max, the error was 0.1 to 0.25 reps when they called “9”, versus 0.7 to 1.2 at “6” (Odgers et al., 2021). In 24 trained men and women on the bench press at 75% of max, it was 0.65 reps on average, over or under (Refalo et al., 2024).

Is RPE just a feeling?

Many think so, but setting your weight by it works. Across 6 studies in trained lifters, adjusting the weight by feel or by bar speed raised the max as much as a fixed percentage of max set in advance. In the 3 RPE studies, the gain was even about 3 kg (about 7 lb) greater, a difference that isn’t statistically clear (Hickmott et al., 2022).

How do you use RPE in the gym?

Rate it at the end of every heavy set of 12 reps or fewer: that’s where the guess is most accurate (Halperin et al., 2022). On very long, very light sets, it can go off the rails: on the leg press, with a very light weight and a cuff restricting blood flow, 20 adults underestimated their reserve by 12 reps on average (Davids et al., 2025).

Then compare it with your target: across 31 studies, the most common use is to aim for a reserve and adjust the weight to match (Bastos et al., 2024). Aimed for 8 and finished at 10? Take a little weight off. No study sets exactly how much.