RIR vs going to failure: how hard should each set be?

Training to failure feels productive. The set that ends with a shaking, grinding last rep seems like the one that builds the most muscle. But when you zoom out from the single set to the whole week, chasing failure on everything usually costs more than it returns. Reps in reserve (RIR) is how you keep effort high without paying that fatigue tax on every set.

What "failure" actually means

Technical failure is the point where you cannot complete another repetition with good form. That is different from feeling like you are done: most lifters, especially newer ones, stop several reps short of true failure while believing they hit the limit. If you want to read more on the underlying scale, our guide on reps in reserve walks through how to estimate it honestly.

RIR simply counts backward from failure. A set left at 2 RIR means you believe you had about two clean reps left. Zero RIR is failure. If you have seen a program prescribe effort as the RPE version of the same scale, it is the same idea counted the other way: an RPE of 8 is 2 RIR. If you are weighing which effort scale to log, the two carry the same reading, and the one you can record most honestly is the one to keep.

Why a small buffer usually wins

Proximity to failure is one of the main drivers of how stimulating a set is for growth, so sets stopped far from failure leave gains on the table. It matters more than the exact rep count, too: a set of 8 and a set of 20 grow muscle about the same when both are close to failure, which is the crux of which rep range to train in. The same holds for tempo: what time under tension actually does for growth is set up hard reps, not stand in for them. Underneath all of it sits one mechanism, and it explains why the load on the bar is not the same as the tension on the muscle. But taking every set all the way to zero has real downsides:

• Fatigue compounds. Failure sets take longer to recover from, which quietly reduces the quality volume you can complete later in the session and the week. Chase failure everywhere for long enough and that buildup is what eventually decides when to take a deload week.
• Form drifts. The last grinding rep is often the ugliest, and on big compounds that is where technique breaks down.
• Estimates get noisy. When every set is a maximal effort, you lose the ability to tell a good day from a bad one.
• It feels more productive than it is. Grinding sets reliably produce soreness the next day, which is easy to read as proof the session worked. It is not: what DOMS actually tells you is how unaccustomed the work was, not how much muscle you built.

This is also why hard sets, not exercises, are the unit worth budgeting across a week. Deciding how many exercises each muscle needs starts from your weekly set target, then fills it with movements.

For most hypertrophy work, keeping the majority of sets around 1 to 3 RIR, with a practical window of roughly 0 to 4 depending on the exercise, captures most of the stimulus at a fraction of the fatigue cost.

Where failure still earns its place

Failure is a tool, not a sin. A few honest sets taken to or very near failure each block are genuinely useful: they recalibrate your sense of effort so your everyday RIR estimates stay accurate. The last set of an isolation exercise is a low-cost place to do this. A drop set on that last isolation movement is a structured way to spend that failure work on purpose, without letting it bleed into the rest of your session. If you want the full case for whether to train to failure every set, including what the research shows and when it is worth the fatigue, we break the decision down separately. The problem is not failure itself: it is failure by default, on everything.

How to log the difference

None of this matters if you do not record it. The value of RIR shows up over time: seeing that a set you planned at 2 RIR actually came in at 4 tells you there is room to add load or reps next session, while a string of 0 RIR sets on a lift is a sign to back off. That effort record is also what tells you when to add weight: hitting the top of your rep range only earns a heavier load if your RIR held while you got there.

RIRLift is built for exactly this. You set an RIR target per exercise on iPhone, then log the load, reps, and actual RIR for every set on Apple Watch as you train: in the moment, when the sensation is still fresh, not reconstructed from memory hours later. Over a block, that history becomes an honest readout of effort you can actually act on. For the practical side of that decision, here is how to run a full session on RIR targets, set by set.