Practical RPE and RIR guide: how to measure training effort without overcomplicating

Last rep of the bench press. You return the bar to the rack and that question stays in your head: did I have two more in me, or was I already at the limit? The answer changes everything. If a lot was left, the stimulus was weak. If nothing was left, you paid a recovery price you may not have needed to pay. And in most gyms, that decision is made on a guess.

RPE and RIR exist to end that guess. They are two ways of measuring the same thing — how much was still in the tank when the set ended — and of turning that reading into a decision: keep the load, raise it or hold back. Without depending on ego, memory or the hunch of whoever was standing next to you.

In this guide you will see the conversion between the two scales, what the science says about training close to failure and a simple way to apply it in your very next session. When you want to take this ruler into the progression decision, the next step is Simple progression: when to add weight.

RPE and RIR: the same ruler, from two angles

RIR stands for reps in reserve: how many good reps you could still do when you decided to stop. RPE stands for rate of perceived exertion: a score for how hard the set was. In lifting, the two meet — RPE 8 means 2 reps were left, RPE 9 means 1 was left, RPE 10 means none were left.

The value lies in what this ruler does to your reading of the workout. Instead of finishing the set thinking "that was heavy", you finish thinking "was it heavy enough to create a stimulus without breaking technique?". That is a better question. And better questions produce better training.

What the science says about training close to failure

The RIR scale used in lifting today was validated in 2016 by Michael Zourdos's team, with two findings that matter for your training: perception gets more accurate the closer to failure you get, and it improves with experience. Translation: you will misjudge the estimate in the first weeks — and that is fine. The ruler calibrates with use.

The second finding frees a lot of people: failing is not mandatory. A 2021 meta-analysis compared training to muscle failure with stopping 1 to 3 reps short and found similar strength and hypertrophy gains in both scenarios. The difference is that whoever stops close to failure accumulates less fatigue — and sustains more quality in the following sets and across the whole week.

That is why the practical recommendation for most working sets sits between RIR 1 and 3. Close enough to failure to recruit the reps that actually stimulate, far enough to protect technique and recovery. Taking everything to the limit is a specific tool, not the week's rule.

How to use it in practice, set by set

The application fits in one honest question at the end of every set: how many good reps were still left? Actually good — with the same range and the same control. A cheated rep to save the ego does not count as reserve.

• Finished with 3 or more left: the set was light for the goal. Note it and consider raising reps or load next time.
• Finished with 2 left (RPE 8): the ideal zone for most of the training. Keep it and consolidate.
• Finished with 1 left (RPE 9): high, productive effort. Use it in the last sets, not in all of them.
• Finished with nothing left (RPE 10): it happens, no drama — but if it becomes routine, recovery sends the bill.

Two precautions make the reading more accurate. First, rate the set right as it ends, not ten minutes later. Second, contextualize the day: bad sleep, stress and a heavy week raise the RPE of the same load — that is not your failure, it is information. A good entry into the session also helps; the warm-up sets say a lot about how the body arrived.

Why this ruler unlocks your progression

Whoever trains without measuring effort swings between two mistakes: raising the load too early and breaking technique, or repeating the same weight for months out of insecurity. RPE/RIR solves both with the same criterion. If the rep range closed with clean technique and there was still margin, there is room to progress. If the set is already flirting with RPE 10, raising the weight now only brings the plateau forward.

This reading talks directly to progressive overload and to double progression: first you push reps within the range, then the load goes up when the top comes with room to spare. The step-by-step for that decision is in Simple progression: when to add weight — and the structure of the week where it happens, in Full body vs. split training.

The mistakes that make the ruler look useless

• Counting a cheated rep as a rep in reserve. If the range shortened, it did not exist.
• Scoring by pride, not by execution. Admitting an RPE 6 hurts the ego, but fixes the training.
• Believing every set needs RPE 9 or 10 to count. The science shows the opposite.
• Ignoring sleep, stress and recovery — the same load changes its score with the day.
• Using the scale for one week and going back to training on autopilot. A ruler without records is a hunch with a technical name.

The last mistake is the most common. Perception only calibrates when you compare what you felt with what happened next — and that requires history. Writing it on paper already works. Having the notes crossed with load, reps and progress works better.

Is it just for advanced athletes? Quite the opposite

Beginners benefit even more. Whoever is starting does not yet know what "close to failure" really is — and the scale teaches that faster than years of trial and error. It avoids the two extremes that most stall beginners: the fear of progressing and enthusiasm without criteria. If you are at that stage, From beginner to results organizes the rest of the base, and How to build more muscle shows where this ruler leads.

For anyone coming back from a break, the ruler works almost like insurance: instead of hunting the old loads in the first week, you let the RIR set the pace of the comeback.

From sensation to data: where SelfShapeAI comes in

Everything in this guide works with a notebook and a pen. The notebook's problem is what it does not do: it does not cross today's perception with the last eight weeks, does not warn you when the same load has been getting easier and does not answer questions mid-workout.

In SelfShapeAI, the session log already asks for load, reps and perceived effort — it takes seconds, between one set and the next. With a few weeks of check-ins, the progress reading shows what memory does not keep: the load that stalled, the muscle group with room to spare on the muscle map, the PRs that arrived without you noticing. And when doubt hits mid-workout — "I closed at RPE 9, do I go up or hold?" — the AI Coach answers on the spot, looking at your plan and your history, not at a generic reply. That difference between AI that follows you and AI that just spits out templates is the theme of Intelligent AI-powered training and the comparison in AI training.

Start in your next session

• Pick two or three main exercises and watch the RIR only on them, so it does not become paranoia.
• At the end of each set, score it honestly: how many good reps were left?
• Log it along with load and reps — without a record, the ruler evaporates by next week.
• After two weeks, compare: are the scores matching the real progress?
• When the top of the range comes with room and clean technique, raise the load. The complete guide is Simple progression: when to add weight.

Frequently asked questions

1. What is RPE in training?It is the set's effort score. In lifting, RPE 8 means 2 good reps were left; RPE 10, that none were left. The full definition is in RPE.
2. What does RIR mean?Reps in reserve: how many reps with good technique you could still do when you decided to stop. It is the most direct way to measure the distance to failure.
3. Do I need to train to failure to build muscle?No. The meta-analysis by Vieira and colleagues (2021) found similar strength and hypertrophy gains training to failure or stopping 1 to 3 reps short — with less accumulated fatigue in the second case. Failure is a specific tool, not a rule.
4. Do RPE and RIR work for beginners?They do, very much. Perception is off more at the start, but calibrates fast when you log and compare. To build the complete base, see From beginner to results.
5. How do I know when to add weight using RPE/RIR?When the top of the rep range comes with stable technique and margin still left (RIR 2 or more) in consecutive sessions, there is room to progress. The complete reasoning is in Simple progression: when to add weight.

Measuring better is training better

RPE and RIR are not jargon to sound technical. They are the simplest way to make every set answer one question: is this taking me where I want to go? When effort becomes data, progression stops being a bet — and training stops depending on guesswork.

In SelfShapeAI, this ruler is already built in: you log the perception on every set and the app turns it into a progress reading and a next step. Create your AI training plan — there is a 14-day free trial to feel the difference of training with criteria.