Should you add weight or reps first when a set gets easier?

The common mistake is adding weight the first session a set feels easier. Add reps first instead. Hold the load until every working set reaches the top of a rep range you chose, then add the smallest step your equipment allows and drop back to the bottom of the range. That answers whether to add weight or reps first for most lifts, most of the time. Two exceptions change the order, and both come down to how big the next step is.

Reps first, then the load

Reps are the smaller step. One more rep at the same load is a change you can make every session, while a plate is a change you can make only when the reps have paid for it. The American College of Sports Medicine puts the trigger in one line: add 2 to 10 per cent to the load once you can do the current weight for one or two reps more than the number you were aiming for [1]. The reps come first in that rule. The load follows them.

The research does not punish either lever for muscle. Plotkin and colleagues gave 43 lifters with at least a year of lower-body training eight weeks of four lower-body exercises, four sets each, twice a week [2]. One group added load at a fixed rep range and the other added reps at a fixed load. Rectus femoris growth modestly favoured the reps group and squat strength slightly favoured the load group, by 2.0 kg, and the authors judged both differences of questionable practical significance [2].

What the order buys is control. Reps first keeps the load fixed while you prove it, and the range stops the reps drifting toward sets of twenty. That pairing is double progression, and it is the Double scheme in the app.

A 70 kg bench press, session by session

Take a lifter benching 70 kg for three working sets, range 6 to 10, every set at RPE 8. RPE is the effort scale where 10 is a true max and 8 means about two reps left. The estimated 1RM column is the best set of each session run through the model behind the one-rep max calculator.

Five sessions at one load moved the estimate from 89.1 kg to 102.9 kg. The weight did not change once. By default the Double rule waits until all three sets reach ten, then adds the 2.5 kg barbell step and resets the target to six reps. Session 6 is where the reset shows its point. Ten reps at RPE 8 implies 80.9 kg for six at the same effort, so 72.5 kg for six should feel well short of RPE 8. Log it at RPE 6 and the estimate reads 98.1 kg, close to where session 5 left it. Log every set at RPE 8 out of habit and the chart shows a drop that never happened.

When the next step is too big for the reps

The first exception is a step that is large next to the load. The app's rule calls any step worth more than 10 per cent of the working weight a chunky jump. On a chunky jump it keeps the load and banks reps past the top of the range first, with a warning that suggests adding reps, adding a set or slowing the tempo before loading up.

A lateral raise shows why. A lifter doing 8 kg dumbbells for twelve at RPE 8 has an estimated 1RM of 12.3 kg. The next pair up is 10 kg, a 2 kg dumbbell step. Eight reps with the 10s at RPE 8 implies 13.5 kg, a bigger jump than the twelfth rep earned. The rule sets a target of thirteen reps at 8 kg instead, which reads as 12.6 kg. The post on how much weight to add each week covers step sizes by equipment.

When the app adds more, or holds

The second exception runs the other way. If every working set beats the top of the range by two or more reps, the load was too light, and the Double rule adds two steps at once and resets to the bottom. On the 70 kg bench above, three sets of twelve would send the lifter to 75 kg for six.

Two things hold the load even when the reps are in. On a cut, a load increase never leaks through: the app keeps the weight and the reps and says that maintaining is a win. And when a block plans its effort week by week, a set logged harder than that week asks keeps the weight where it is until the same reps come easier. Neither one is a stall. A stall is three sessions without progress, and the post on when a lift stalls covers that case.

Is strength the exception?

Partly. If the goal is a heavier one-rep max rather than bigger muscles, load matters more than the order. A meta-analysis of 21 studies with every set taken to failure found 1RM gains favoured loads above 60 per cent of 1RM, while muscle growth was similar across loads [3]. A strength lifter still adds reps first inside a range. The range is just lower, such as 3 to 5, so the load climbs more often.

Questions

How many reps over the target before adding weight?

One or two, on every working set. The ACSM position stand puts the trigger at one or two reps over the target number, followed by a load increase of 2 to 10 per cent [1]. In the app's Double scheme the default trigger is every working set reaching the top of the range. Beat the top by two or more on every set and it adds two steps instead of one.

Should I add a rep to every set or just the first?

Add reps wherever they come, but wait for all of them before adding load. In the bench example, the first set reached ten a session before the last one did. The app's default waits for every working set. You can switch the trigger to your best set or to the average, which moves the load up sooner on lifts where fatigue drags the last set down.

Does adding reps build as much muscle as adding weight?

Close enough that the choice should come down to the equipment. In 43 trained lifters over eight weeks, progressing reps at a fixed load and progressing load at a fixed rep range produced differences in growth and strength that the authors judged of questionable practical significance [2]. Pick whichever lever your next step allows, and keep the effort honest.