Journal · · 9 min read · Updated
Double progression explained: add reps until the range runs out, then add weight
Linear progression asks the bar to get heavier every session, and on most lifts that stops working within months. What usually comes next is double progression: the weight stays put while the reps climb through a range, and only then does the load go up. It is simple to describe and easy to run badly. This post covers the rules, eight sessions worked through, what a small plate adds in percent, and the signs it has run its course.
Most strength programs start with linear progression: a little more weight every session. On most lifts it works for a few months and then stops, usually on the press first. How to tell when that phase is over is in choosing a strength program you will finish. What most people run next is double progression. It is also one of the progression schemes in the strength log I build, so this is the long version.
What is double progression?
You progress two things in a fixed order: reps first, at a weight you keep, then load, once the reps have run out of room.
| Part of the rule | What it means |
|---|---|
| Rep range | A bracket for every working set, written sets × bottom-top: 3 × 6-10 |
| Hold | Keep the weight while any set is short of the top of the range |
| Add load | When every working set reaches the top, add the smallest increment you have |
| Reset | At the new weight, expect to be back near the bottom of the range |
| Bad session | A set below the bottom is a miss. Repeat the weight. Only several misses in a row call for a lighter restart |
None of this is new. The American College of Sports Medicine’s 2009 position stand on progression lists increasing load within a prescribed zone, its example being 8-12 RM, as one of three ways to progress. Its rule for when to move is a 2-10% increase once you can do one or two reps more than the target on two consecutive sessions, at the lower end for small-muscle exercises (ACSM, 2009). Double progression is that rule with the bookkeeping written down.
A worked example: eight sessions of overhead press
Barbell overhead press, 3 × 6-10, starting at 40 kg. The smallest jump in this gym is 2.5 kg.
| Session | Load | Reps (set 1 / 2 / 3) | Total reps | Next session |
|---|---|---|---|---|
| 1 | 40 kg | 10 / 8 / 7 | 25 | Same weight |
| 2 | 40 kg | 10 / 9 / 8 | 27 | Same weight |
| 3 | 40 kg | 10 / 10 / 8 | 28 | Same weight |
| 4 | 40 kg | 10 / 10 / 10 | 30 | Add 2.5 kg |
| 5 | 42.5 kg | 8 / 7 / 6 | 21 | Same weight |
| 6 | 42.5 kg | 8 / 7 / 5 | 20 | Same weight (set 3 missed) |
| 7 | 42.5 kg | 9 / 8 / 7 | 24 | Same weight |
| 8 | 42.5 kg | 10 / 9 / 8 | 27 | Same weight |
Sessions 1 to 4 are the easy part: the weight never changes while the total climbs from 25 reps to 30, and session 4 is the first time all three sets reach ten.
Session 5 is where people lose faith. Going from 10, 10, 10 to 8, 7, 6 looks like a step backwards. It is the plan working: every set is inside the range at a heavier weight, and the climb starts again.
Session 6 has a set below six. That is a miss, and the response is nothing special: same weight next time. A rough session right after a 6% jump is common.
By session 8 the lifter is doing 27 reps at 42.5 kg. Reps times load has gone from 1,000 kg in session 1 to about 1,148 kg, roughly 15% more work, while the bar moved once. The load moves less often, and the reps show you are still progressing between jumps.
Which add-weight rule should you use?
“Every set reaches the top” is the strict version and the most common one. There are others.
| Variant | Add load when | Suits | Catch |
|---|---|---|---|
| All sets at the top | Every working set reaches the top of the range | Most lifts; the safe default | Slowest of the single-session rules. One tired last set holds the lift back |
| Two sessions in a row | All sets reach the top in two consecutive sessions | Lifts where one good day can be luck | Slower again. Worth it mainly on lifts close to your ceiling |
| Rep total | The session total reaches a target, say 27 across 3 × 6-10 | Lifts where the last set always lags | Lets one set drift low while the others carry it |
| Top set only | The first, heaviest set reaches the top | A top set followed by lighter back-off sets | Moves load fast; the back-off sets need their own rule |
Pick one per lift and keep it for the block. Switching rules midway is how a lift ends up at a weight no rule would have chosen.
Why does adding reps count as progress?
Because, over a couple of months at least, it seems to work about as well as adding load. In Plotkin et al. (2022), 38 trained lifters did eight weeks of lower-body training. One group added load and kept its reps in a range; the other kept its load and added reps. Squat strength rose 21.8 kg in the load group and 19.3 kg in the reps group, a gap the authors called of questionable practical significance, and muscle thickness changes were similar.
The wider rep-range research points the same way. A 2017 meta-analysis led by Brad Schoenfeld found similar muscle growth with light and heavy loads when sets were taken to failure, while one-rep max strength favoured heavier loads. A 2021 re-examination of the repetition continuum, also led by Schoenfeld, argued that adaptations happen across a wide range of loading zones. The ACSM’s 2026 update, an overview of 137 systematic reviews, found hypertrophy did not differ across 30-100% of one-rep max when effort was comparable, and that strength gains were greater at 80% and above (Currier et al., 2026).
So no rep range is magic. For size, pick a range you will push honestly. For a lift where top-end strength is the point, keep the range near the heavy end, where the strength benefit shows up.
How wide should the rep range be?
Wide enough that one load jump does not throw you below the bottom of it.
| Range | Typical use | What to watch |
|---|---|---|
| 3-5 | Main barbell lifts, strength focus | Narrow. One standard jump can eat the whole range, so microload |
| 6-10 | Compound lifts, mixed goals | Room for a 4-6% jump to cost two or three reps |
| 8-12 | Most machines, dumbbells, general size work | The zone ACSM’s 2009 stand used as its example |
| 12-20 | Small-muscle isolation: raises, curls, calves | The next dumbbell is often a 15-25% jump, and the range has to absorb it |
If you add weight and land below the bottom of the range two or three sessions running, the range is too narrow for the jump. Widen it, or make the jump smaller.
What is microloading, and when do you need it?
Microloading means adding less than the standard 2.5 kg (5 lb) step, with fractional plates from 0.25 to 1.25 kg or small add-on weights for dumbbells. The case for it is arithmetic. The same plate is a very different jump on different lifts:
| Working weight | +0.5 kg | +1 kg | +2.5 kg | +5 kg |
|---|---|---|---|---|
| 30 kg overhead press | 1.7% | 3.3% | 8.3% | 16.7% |
| 60 kg bench press | 0.8% | 1.7% | 4.2% | 8.3% |
| 100 kg squat | 0.5% | 1.0% | 2.5% | 5.0% |
Against ACSM’s 2-10% band, which asks for the low end on small-muscle exercises, 2.5 kg on a 30 kg press sits near the top of a range meant for big lifts. The same plate on a 100 kg squat is near the bottom. That is a large part of why the press stalls first.
Dumbbells are harder, because the rack decides the jump. Going from a 10 kg to a 12 kg dumbbell for lateral raises adds 20%. Without add-on plates the practical fix is a wider range, such as 12-20, so the reps can absorb the jump.
You need microloading when the smallest standard jump keeps knocking you out of the range: overhead presses, bench press for lighter lifters, most single-joint lifts. You can usually skip it on squats and deadlifts until the weights are well past bodyweight.
When does double progression stop working?
On accessories it may never need replacing. On main lifts, these are the signs I would look for before changing:
- Reps flat at the same weight for three or four weeks, on more than one lift at once. One stuck lift is a problem with that lift. Several at once is usually recovery, food or sleep, and no progression model fixes those.
- Every jump lands you below the range, even after microloading. The lift may need a heavier, narrower range for a while, or a different variation.
- You want to peak a main lift. Double progression is good at building and has no plan for a heavy single on a set date. A percentage cycle off a training max, such as 5/3/1, does, starting from a one-rep max that is tested or estimated.
Before giving up on the model, change one thing inside it: swap the exercise variation, move the range (from 6-10 to 10-15 for a block), or add a set. If one lift does move to a percentage cycle, the others can keep running double progression. Mixing models across lifts is normal.
This is training information, not medical advice. If a load jump hurts rather than just feels hard, have a clinician or a coach who can watch you lift take a look.
How LiftLog runs it
In LiftLog, progression is a setting on each exercise in a program rather than on the program as a whole. The schemes are Double progression, Linear, % of training max, and RPE / RIR.
With a scheme on an exercise, each pending set arrives filled in with the weight the program prescribes, under a chip such as “Target 80 kg × 8 · Double”. When the workout ends, Workout Complete includes a progression summary, and the Progression review labels each exercise Progressed, Hold, Missed, or Deload. Smart weight suggestions are on by default; one switch in Settings turns them off if you would rather keep the books yourself.
For microloading, the Plate Calculator shows which plates go on each side of the bar and notes when a weight can’t be loaded exactly. If you own fractional plates, check the suggested weight against what is actually on your rack.
Questions
Frequently asked
What is double progression in weight training?
When should I increase the weight in double progression?
What rep range should I use for double progression?
What is microloading?
Is adding reps as effective as adding weight?
When should I stop using double progression on a lift?
Mentioned in this post
Apps in this post
More from the journal