Too easy or too hard? When to change the weight mid-workout
Too easy, too hard, missed reps or form breaking down: what to do on your very next set, with kg and lb numbers, and why falling reps aren't a reason to drop.
Short answer
Most programs give you a weight before you’ve lifted it. The set itself tells you whether that weight was right for today. You don’t need to wait until next week to act on it: you can change the very next set.
The tool for that is reps in reserve: how many more clean reps you could have done when you racked it. Rate it right after each working set, before you rest.
What to do on the next set
| How the set went | Next set |
|---|---|
| Too easy: 4 or more reps left | Add your smallest jump: about 2.5 kg (5 lb) on a barbell, or the next dumbbell up |
| On target: 1 to 3 reps left | Keep the weight |
| Very hard, but you got every rep | Keep the weight and expect a rep fewer next set |
| Fell short of your rep range | Take off about 5 to 10% |
| Missed the last rep | Rest longer and take off 5 to 10%. Retry the same weight once only if you can name and fix a setup mistake |
| Form broke down to finish | Take off 5 to 10% and own every rep |
| Pain, especially sharp or in a joint | Stop that exercise. This isn’t a weight decision |
Two numbers to keep in your head:
- Going up: one small jump at a time. On a 60 kg bench press, 2.5 kg is about 4%. On 10 kg dumbbells, the next pair up can be 20% more, so expect fewer reps after the jump.
- Going down: 5 to 10%. On a 100 kg squat, that’s 5 to 10 kg. Smaller cuts often leave you in the same spot.
Falling reps are normal
Say you get 10 reps on your first set and 8 on your second at the same weight. That doesn’t mean the weight was wrong. Fatigue builds across sets.
In a review of 29 studies where people did repeated hard sets at a fixed load, set 2 averaged about 70% of the reps of set 1, set 3 about 55%, and sets 4 to 6 about 45 to 50% (Nuzzo, 2024). Longer rest between sets made the drop smaller.
Those sets were taken to failure. If you’re leaving 1 to 3 reps in reserve, your drop will probably be smaller, though that’s our extrapolation, not something the review measured. Either way, the rule is the same:
- Lost a rep or two but stayed in your range? Keep the weight.
- Fell below the bottom of your range? Take some weight off.
- Need more reps on later sets? Rest a little longer before you cut the load.
Most people guess their reps left too low
A review of studies on this found lifters underpredicted how many reps they had to failure by about one rep on average (Halperin and colleagues, 2022). That’s an average across studies: individual guesses missed in both directions. Guesses got slightly more accurate closer to failure, and training experience didn’t seem to help much.
What that means in practice:
- Treat your reps-left number as an estimate, not a measurement.
- If a set felt like it had 4 or more left, it was probably too light. Go up.
- To calibrate, take the last set of an exercise to failure now and then, but only where failing is safe: on a machine, or with safety arms set.
The RPE scale and reps in reserve describe the same thing. RPE 8 means about 2 reps left.
When you fail a rep
Failing a rep isn’t a disaster, but it changes the plan for the next set.
- Rack it safely. Use safety arms or a spotter on squats and bench.
- Rest longer than usual, an extra minute or so.
- By default, take off 5 to 10% for the rest of your sets. A miss usually means you were at your limit today.
- Retry the same weight once only if you can name a specific setup mistake, like your feet or grip being off, and fix it. “My form slipped” isn’t a setup mistake; that’s fatigue, so go lighter.
One bad set doesn’t mean a bad day. Two misses in a row on the same exercise usually does. Finish your sets lighter and move on.
Going up mid-workout
If your first working set leaves 4 or more reps in reserve, the weight is too light today. Add your smallest jump and judge the next set the same way. Don’t add two jumps at once unless the set was very easy, such as 6 or more reps left.
Strength research supports this kind of adjustment. Across 8 small studies, adjusting the load to performance on the day improved maximal strength more than fixed loads did (Zhang and colleagues, 2021). The studies were small, so treat that as a reason to try it, not a promise.
Pain is not a load decision
If an exercise hurts, especially a sharp pain or pain in a joint, stop that exercise. Don’t just lighten it and push on. Swap to something that doesn’t hurt, or end that part of the session. If pain keeps coming back, get it checked by a qualified professional.
Next session
Mid-workout changes fix today. To decide next session’s weight, look at your whole session. If every set reached the top of your rep range with a rep or two to spare, add weight next time. If the last set was a grind, repeat the weight once. If not every set reached the top, keep the weight and add reps. The double progression calculator applies the same rules and gives you a next-set call from the reps you enter. It assumes one weight for every set, so if you changed weight mid-session, enter the sets at the weight you’ll start with next time.
Founder and editorial lead, Handler
Handler's founder writes general training guides and documents their sources. This article is educational, not an individual coaching or medical assessment.
Sources
The study findings below come from the linked abstracts. The decision table and the 5 to 10% and smallest-jump rules are coaching guidance, not results from those studies.
Sources
- 01Halperin et al. 2022, Sports Med: accuracy in predicting repetitions to failure (Lifters underpredicted reps to failure by 0.95 reps on average; accuracy improved closer to failure; training status didn't seem to matter.)pubmed.ncbi.nlm.nih.gov/34542869/
- 02Nuzzo 2024, J Strength Cond Res: repetitions across repeated fatiguing sets (At a fixed load, sets 2 to 6 reached about 70, 55, 50, 45 and 45% of set-1 reps; longer rest reduced the drop.)pubmed.ncbi.nlm.nih.gov/38781472/
- 03Helms et al. 2016, Strength Cond J: the reps-in-reserve RPE scale (A reps-in-reserve scale lets you adjust load set by set to match the day.)ro.ecu.edu.au/ecuworkspost2013/2114
- 04Zhang et al. 2021, Front Physiol: autoregulation vs fixed loading (Across 8 small studies, autoregulated loading improved maximal strength more than fixed loading.)frontiersin.org/articles/10.3389/fphys.2021.651112/full