RPE and reps in reserve: rate effort like a coach

The plan says 80% for five. On Thursday it's a grind and the last rep takes three seconds. Ten days later, same bar, same five reps, and it moves like a warm-up. Nothing in the spreadsheet accounts for that.

Rating effort does. RPE, rating of perceived exertion, is how hard a set felt on a 1 to 10 scale. Reps in reserve, RIR, is how many clean reps you had left when you racked it. On the scale lifters use, they're the same number said two ways. RPE 8 means 2 in reserve, RPE 9 means 1, RPE 10 means nothing left.

That's the easy part. The hard part is that most people read their own scale wrong by about two reps, and new lifters by more.

The scale is a rep count, not a feeling

RPE started in cardio research as a way to rate how hard your breathing and legs felt. Lifting borrowed it and made it concrete. Instead of guessing at an abstract effort score, every number on the lifting version is pinned to a count of reps you could still have done. Zourdos and colleagues validated that version in 2016 by checking lifters' ratings against how fast the bar actually moved, and it's been the standard in strength coaching since.¹

RPE	Reps in reserve	Where it belongs
10	0, nothing left	Testing, and a few times a year to reset your scale
9	1	A last set, or a lift you can fail safely
8	2	Most working sets
7	3	Most working sets
6 and under	4 or more	Warm-ups and technique work

RIR is the more literal of the two, and for most people the easier one to answer. "How many more could I have made" is a question with a number in it. "How hard was that out of ten" is a question about a feeling.

Why not just use percentages?

Percentage programming looks precise. Test your max, take 75% of it, write it down. The trouble is that the 75% is anchored to one particular Tuesday, and your real max moves.

A short night, a stressful week, a skipped meal, a couple of years since the test: all of it changes how much force you can produce today. So 75% can land at a genuine RPE 7 on a good day and an RPE 9 on a flat one. You prescribed a load and your body received an effort, and it wasn't the one you meant to give it.

A percentage locks the weight and lets the effort wander. Rating effort locks the effort and lets the weight move, which is usually what you actually care about. That's the case the coaches who formalized RIR-based programming made about a decade ago.² On a strong day you add a little; on a flat day you take some off. The stimulus lands where you wanted it either way.

Most lifters are off by about two reps

A 2018 study in the Journal of Strength and Conditioning Research asked 48 trained adults to call how many reps they had left during sets at 70 to 80% of their max. They were off by roughly 2 reps on chest press and about 3 on leg press, and the error was worst on the first set of an exercise before shrinking on later sets.³

Experience helps but doesn't fix it. Later work put trained lifters at about 2 reps of underprediction on average, with accuracy climbing as training age climbs and never quite reaching perfect.⁴

The direction of the error is the useful bit. People stop short. They feel close to failure and they aren't. So a beginner who says "that was everything I had" often has 3 or 4 clean reps left, and their ratings run scared: sets get called harder than they were. The scale is only as good as the person reading it, and a rating is an estimate. Treat it like one.

Building a scale you can trust

Calibration is a skill you build on purpose, and three habits do most of the work.

The first is taking a set to true failure now and then. Not heavy, not uncomfortable: you go for one more and it doesn't move with decent form. Do that a handful of times across your main lifts and you have a felt reference for what a real 10 is. Every 7 and 8 you call afterwards gets measured against that memory. With no real 10 in your body, the whole scale floats.

The second is watching the bar. In that validation work, average bar speed tracked effort closely. As sets got harder the bar slowed, and the link was strong, about r = -0.88 in experienced lifters and -0.77 in novices.¹ You can see this on your own phone. The last hard rep before failure crawls compared to the first, and learning to read that slowdown gives your rating something outside your own head to hang on.

The third is checking yourself. Every so often, push a set two reps past where you'd have stopped and find out whether you were right. Over a few months the gap between what you call and what's true tends to close.

Most sets belong at 2 or 3 in reserve

You don't want to be near failure on every set. The target that keeps showing up in strength research is 1 to 3 reps in reserve, RPE 7 to 9.

Almost all of the benefit lives just short of failure, while the cost of going all the way climbs fast. Grinding to a true 10 every set piles on fatigue, wrecks the sets that follow, and is where form falls apart, which for most people is the real risk. Leaving a rep or two in the tank keeps quality high, so you can do more good sets across the week. That weekly total is what drives the result, and it gets its own article: how many sets per week your muscles actually need.

A workable default: keep most working sets at 2 to 3 RIR, and dip toward 1 RIR only on a last set, or on a movement you can fail safely, like a leg press. Save the all-out efforts for testing, not for every Tuesday.

Where the scale gets slippery

Ratings are subjective, and in new lifters they're often badly off, which is the whole reason the accuracy studies exist. A beginner's RPE 9 and a coach's RPE 9 can be two full reps apart. Ratings also drift with mood, caffeine, and how much you slept, so a set that felt like an 8 in March can get called a 9 in November for no real reason.

Bar velocity is harder to fool, which is why velocity-based training exists. A sensor reading how much speed you lose within a set doesn't talk itself into or out of how hard the set was. The trade is that reading velocity well takes a tracking device or a trained eye.

None of it settles your case specifically, either. Effect sizes vary a lot from person to person, and most of these studies ran on younger, trained, mostly male subjects. Treat the numbers as starting points.

Common questions

Is RPE or reps in reserve better for beginners? Reps in reserve, usually. "How many more could I have done" is a more concrete question than "how hard was that." Expect your estimates to be off at first, and lean toward assuming you had more left than it felt like.

What RPE should I train at to build muscle? Most working sets in the RPE 7 to 9 range, which is 1 to 3 reps in reserve. That covers the large majority of muscle and strength gains. Hitting failure on everything mostly adds fatigue and form risk for very little extra.

Can I use RPE if I train at home without much equipment? Yes. RPE and RIR measure your effort, not the load on a specific machine, so they work the same with dumbbells, bands, or bodyweight as they do with a barbell.

The set you can't watch yourself

In those accuracy studies the error was biggest on the first set and biggest in the least experienced lifters, and it shrank steadily with practice. That tells you who's already covered. Anyone a few years into training who keeps an honest log and takes the odd set to genuine failure is running their own calibration loop, and a training partner with a good eye does most of the rest for free.

Early on is different. Your internal scale is least reliable exactly when you have the least reference for it, and you can't see your own bar slow down from underneath it. A coach watching live is an external RPE meter. They read the same slowdown the velocity work measures, catch the set where your RPE 8 was really a 6, and change the load before you bank an easy set or a sloppy one. It's the same reason a coach watches every rep instead of reading it off a log the next day, and you can see how that runs on our how it works page. Founding membership is $49 your first month, then $299/mo locked for life, with 25 spots. What you're paying for is an eye trained to read effort you can't feel yet.

Sources

1. Zourdos and colleagues, 2016. Novel Resistance Training-Specific Rating of Perceived Exertion Scale Measuring Repetitions in Reserve. Journal of Strength and Conditioning Research. ↩ ↩²

2. Helms and colleagues, 2016. Application of the Repetitions in Reserve-Based Rating of Perceived Exertion Scale for Resistance Training. Strength and Conditioning Journal. ↩

3. Hackett and colleagues, 2018. Estimation of Repetitions to Failure for Monitoring Resistance Exercise Intensity: Building a Case for Application. Journal of Strength and Conditioning Research. ↩

4. Armes and colleagues, 2020. Just One More Rep: Ability to Predict Proximity to Task Failure in Resistance Trained Persons. Frontiers in Psychology. ↩