RPE — Rate of Perceived Exertion

What RPE actually measures

RPE — Rate of Perceived Exertion — is a 1-10 self-rating of effort during a set. The concept originated in cardiovascular research with Gunnar Borg’s 6-20 scale in the 1960s, and was later simplified to a 1-10 version that mapped more cleanly onto subjective effort. The strength-training adaptation belongs to Mike Tuchscherer, whose Reactive Training Systems work in the early 2000s reframed the scale around reps in reserve rather than vague feelings of difficulty.

The point is autoregulation. Your real strength varies day to day, often by 5 to 10 percent, based on sleep, food, stress, hydration, and the time of day you train. A fixed percentage of 1RM does not account for any of that. RPE does. Instead of asking how heavy the bar is in absolute terms, it asks how hard the set actually felt — and then trusts that answer to set the load.

The scale

RPE 10 — Maximum effort. You could not have done another rep.
RPE 9.5 — Maybe one more rep, maybe not.
RPE 9 — One more rep in the tank.
RPE 8.5 — One to two reps in the tank.
RPE 8 — Two more reps in the tank.
RPE 7 — Three more reps in the tank.
RPE 6 — Four more reps. Speed work range.
RPE 5 — Easy. Warmup territory.
RPE 4 and below — Below working weight.

The most-used RPEs in serious training are 7 through 10 for working sets, and 5 through 6 for warmups. Below 5 is bar work, mobility, and the lightest plates of a ramp. Above 9 is reserved for testing days, last sets of a peaking block, and the occasional honest top single.

RPE-to-RIR conversion

• RPE 10 = 0 reps in reserve (RIR 0)
• RPE 9 = 1 RIR
• RPE 8 = 2 RIR
• RPE 7 = 3 RIR
• RPE 6 = 4 RIR

RIR (Reps in Reserve) is the inverse of RPE — same idea, opposite labels. Different programs prefer different framings. RTS-influenced powerlifting plans speak in RPE; modern hypertrophy programs from Mike Israetel and Eric Helms tend to use RIR. The math is the same. Pick the language that helps you make honest decisions in the rack.

Why RPE works

Three reasons, in increasing order of importance.

Real strength is not constant. Your true 1RM on a given day fluctuates roughly plus or minus 10 percent based on sleep, food, stress, hydration, and circadian rhythm. A program that prescribes “5x3 at 85 percent” is asking for 85 percent of yesterday’s 1RM, not today’s. On a tired day that load is RPE 10 and overshoots the intended difficulty; on a fresh day it might be RPE 7.5 and undershoot it. Either way, the session does not train what the program designed.

Effort matters more than load. Decades of training research, from Zatsiorsky onward, point in the same direction: a set taken to a specific proximity to failure produces a similar adaptation across a meaningful range of loads. A set at RPE 8 trains hard enough to drive strength and hypertrophy without burying the lifter in fatigue. The exact kilograms on the bar are downstream of that effort target.

It is self-corrective. Bad sleep night? RPE 8 today is automatically lighter than RPE 8 last week. The program adapts to the lifter, not the other way around. Over a year of training, this is the difference between a plan that survives real life — busy work weeks, travel, sickness, family — and a plan that quietly grinds the lifter into the ground.

RPE and percentages — using both

Most modern programs use a hybrid: percentages prescribe the warmup ramp, RPE prescribes the working set difficulty. The percentages give structure; the RPE absorbs the day’s variability.

Example for a squat top set:

Warm-up:
- Bar x 10 (RPE 4)
- 50% x 5 (RPE 5)
- 70% x 3 (RPE 6)
- 85% x 1 (RPE 7)
- 90% x 1 (RPE 8 — first feeler single)

Working sets:
- 5x3 @ RPE 8 — load adjusts based on how the warmup ramp felt

If the warmup felt heavy — the 70 percent triple already brushing RPE 7 — your working weight on the 5x3 is lighter than usual. If it felt unusually easy, heavier. The feeler single at 90 percent is the deciding data point: it tells you whether today’s RPE 8 lives at 92.5 percent or at 87 percent. Either is fine; both train the intended effort.

A worked example

An intermediate lifter with an estimated 1RM squat of 150 kg. Today’s program is 4x4 at RPE 8.

Tuesday — slept eight hours, ate well, no work stress:

• Warm-up: 60 kg x 5 (RPE 5), 100 kg x 3 (RPE 6), 125 kg x 2 (RPE 7)
• 4x4 at RPE 8 lands at 130 to 132.5 kg

Saturday — slept five hours, stressed at work, ate light:

• Same warmup ramp feels heavier. 100 kg x 3 already feels like RPE 7.
• 4x4 at RPE 8 lands at 122.5 kg

Same RPE, different absolute load — and that is the point. Saturday’s 122.5 kg trained the same physiological adaptation as Tuesday’s 130 kg, because both sessions reached the same effort level. Forcing 130 kg on Saturday would have produced one of two outcomes: missed reps, or a session that drained the next two days of training. Neither is worth the false consistency on the spreadsheet.

Common mistakes with RPE

Under-rating effort. Beginners systematically rate sets one to two points lower than they are. A set at RPE 8 (2 RIR) feels like RPE 6 to a new lifter because they have never seen what RPE 10 actually looks like. The rating gets calibrated by occasionally pushing to RPE 10 — once a month, take a top set to true failure on a safe lift like the leg press, the bench (with safeties or a spotter), or a paused front squat. Not the deadlift.

Over-rating effort. Some lifters rate every set at RPE 9 or 10 because it felt hard, even when bar speed was strong and form was clean. The fix is to video the set: if you would have hit two more reps clearly, that was an RPE 8 regardless of how it felt in the moment. Effort is not the same as discomfort.

Using RPE on the deadlift recklessly. Deadlift RPE jumps from 9 to 10 with very little warning. The lift’s bar speed slows late, not early, and the position can deteriorate in a single rep. Cap deadlift working sets at RPE 8 unless you are peaking for a meet, and even then keep the high-RPE work sparse.

Ignoring RPE drift across a session. Set 1 at RPE 8 and Set 5 at RPE 8 are not the same load — fatigue compounds across a session. Modern programming expects load drops of 5 to 10 percent across multi-set RPE work. If your fifth set matches your first, either you sandbagged the first set or you are now grinding past the intended effort.

Using RPE without anchoring. RPE is a rating, not a number. You need to know your true 1RM, or a recent honest estimate of it, to interpret what RPE 8 should feel like. Reference the one-rep-max-calculator to estimate yours from training data, and revisit it every six to eight weeks.

When NOT to use RPE

There are situations where RPE is the wrong tool, and a percentage-based or rep-target prescription serves better.

• First six months of training. Linear progression on a beginner program is faster, and your RPE accuracy is too low to be useful. Add weight, log the set, repeat.
• Final week before a powerlifting meet. Use percentages — the singles need to be exactly the weight, not approximately RPE 9. Peak week is about confirming planned attempts, not autoregulating.
• High-rep accessory work. RPE 8 on a bench press at 12 reps is harder to estimate than at 4 reps, because the relationship between reps left and effort gets noisy in long sets. Use rep targets and RIR (for example, 3x12 with 2 RIR) instead.

RPE in different rep ranges

• 1 to 3 reps: classic RPE territory. Most useful here, because the gap between RPE 8 and RPE 10 maps cleanly to one or two reps.
• 4 to 6 reps: still very useful. Standard hypertrophy-strength work. Bar speed is the easiest cue.
• 8 to 12 reps: useful for bodybuilders, but switch to RIR framing if you find yourself second-guessing the number.
• 15 reps and above: RPE accuracy drops sharply. Use rep targets with form failure as the cap — for example, “3 sets to technical failure” rather than “3x20 at RPE 8”.

How to track RPE in Carve Log

The workout logger lets you note RPE per set. Over time, the average RPE for a given load tells you how the lift is trending: a load that used to be RPE 8 dropping to RPE 7 over four weeks is unambiguous progress. Coupled with the one-rep-max-calculator, you can confirm whether your estimated 1RM is rising or stalling. Reference progressive-overload for using RPE as a progression lever, and rest-between-sets for managing the rest periods that make RPE ratings meaningful in the first place — rushed rest inflates RPE artificially.

Final word

RPE is not a religion. It is a tool that handles the irreducible variability of training. Anchor it to real loads, calibrate it with occasional max-effort sets, and you will be programming your weeks around effort rather than rigid percentages — which is how strong, healthy lifters train for decades. The spreadsheet is a model of training; the RPE rating is the lifter telling the spreadsheet what actually happened in the rack today. When the two disagree, the lifter wins.