Heavy or light? Rep ranges for muscle and strength

For muscle, heavy and light both work. Sets of 5 and sets of 15 build about the same amount of muscle, as long as each set is hard.^[1][2] For a bigger one-rep max, heavier sets win: about 80% of your max or more, which works out to roughly 8 reps or fewer.^[2]

So most people do well with mostly moderate sets, plus some heavier sets on the lifts they want to get stronger at. That last part is coaching practice, not a tested finding.

Why isn't this obvious? Because gym advice likes neat boxes: heavy for strength, moderate for size, light for endurance. Only the strength box holds up well.^[3] In how hard a set should be, you learned to judge effort. This article covers the other dial: how heavy the weight is.

Why it matters

A few terms first. Load means how heavy the weight is compared with what you can lift. Researchers write it as a percentage of your one-rep max (1RM): the most weight you can lift once with good form. If your max is 100 kg, 80% is 80 kg. A rep range is the span of reps you aim for in a set, such as 8–12. The heavier the load, the fewer reps you can do.

Get load wrong and you lose time in one of two ways.

Some beginners believe only heavy weights build muscle. So they load the bar before they can control it. A position stand is an official summary of the evidence by a group of experts. The American College of Sports Medicine (ACSM) published one in 2026. It notes that people without experience may need to start with low loads, partly for safety.^[1]

Others hear "light weights work" and take it too far. They do easy sets of 15 with a weight they could lift 30 times. Failure is the point where you can't complete another rep. In one trial, light sets that stopped well short of failure grew far less muscle than light sets that went all the way.^[4] Light weights work. Easy sets don't.

What the research says

Some more terms. A systematic review collects every study on a question using fixed rules. A meta-analysis pools the results of many studies on the same question into one estimate.

Muscle grows about the same across a wide range of loads

ACSM's 2026 position stand pulled together 137 systematic reviews covering more than 30,000 people. Load didn't change how much muscle people built, from light loads around 30% of max up to the heaviest, 100%.^[1] Much of that evidence came from people new to training, like you.^[1]

Strong evidence. The research is consistent here.

The pattern also held in people who already lifted. In one trial, 49 trained men did either 20–25 reps or 8–12 reps per set for 12 weeks, every set to failure. Both groups gained a similar amount of muscle. Strength went up in both. Only the bench press differed: its max went up about 14 kg in the heavier group and 9 kg in the lighter one.^[7]

In another trial, 18 trained men did either 25–35 reps or 8–12 reps for 8 weeks. Muscle grew about equally. But the squat max rose 19.6% with the heavier sets and 8.8% with the lighter ones. The light sets improved muscular endurance more, measured as reps at a light weight.^[8]

Limited evidence. These are two small trials, and both used only men. They point the same way as the meta-analyses.

Light only works when the set is hard

One trial took 25 untrained men through 8 weeks of leg extensions. Light sets at 30% of max, taken to failure, grew the thigh 7.8%. Heavy sets at 80% grew it about 8%, whether or not they went to failure. Light sets that stopped well short of failure grew it 2.8%, a change too small for the study to be sure of.^[4] Limited evidence. It's one exercise in one small group of men.

ACSM's main recommendation is to train with high effort, and you can reach that with many different loads. It proposes that stopping about 2–3 reps short of failure is enough effort.^[1] Expert consensus. This is the panel's view, not a tested finding.

There's a floor, somewhere around 30%

The lightest weights seem to fall short. In one trial, 30 men trained one arm and one leg at 20% of max, and the other side at 40%, 60% or 80%. The number of sets on the heavier side was adjusted so the total weight lifted matched.

After 12 weeks, the 20% side grew about half as much. On the thigh, that was 8.9% against about 20%. The 40%, 60% and 80% sides grew about the same.^[9] Limited evidence. It's one trial in men.

Review authors put the floor somewhere around 30% of max. Above it, they write, growth is similar across loads.^[3] Expert consensus. This is the authors' view, not a tested finding.

Heavy wins for strength, partly because it's practice

ACSM says load affects strength more than any other training result, and it lists 80% of max or more for strength.^[1] Strong evidence, with the 2023 meta-analysis above pointing the same way.^[2]

Part of the reason is practice. A 1RM test is a heavy lift, and heavy training rehearses it.

Review authors point out that strength tests are usually done on the same exercises people trained with. An isometric test is different: a machine measures your force without you lifting a weight. On that kind of test, they say there's little difference between heavy and light training.^[3] The 2017 meta-analysis found the same: no clear difference in isometric strength.^[5] Limited evidence that the 1RM gap is partly practice. A newer analysis, below, couldn't settle it.

A 2026 meta-analysis of 10 studies, with 245 people, looked at strength tests unlike the training itself. The results leaned towards heavy loads, but the studies can't tell it apart from no effect.^[10] Mixed evidence on whether heavy loads also build more strength you can use elsewhere.

ACSM's advice follows from this: train in a way that matches the result you want.^[1] If you want a heavier squat, some of your squats should be heavy.

Why most people land on moderate loads

If light and heavy both build muscle, why not just pick one? Review authors give practical reasons.

Light sets mean many more reps, so sessions take longer. They also cause more discomfort, which the authors think could make people less likely to stick with training. Heavy loads need more sets than moderate ones for the same growth, and heavy loads with many sets put more stress on the joints.^[3]

So the authors lean towards moderate loads for most muscle work.^[3] Expert consensus. This is the authors' view, not a tested finding. How many sets you need is its own topic: see how many sets per week build muscle.

The same authors say mixing rep ranges might help growth, but they call the practical value questionable and say more study is needed.^[3] Limited evidence.

Your reps at a given weight are your own

A 2024 meta-analysis pooled 269 studies with 7,289 people. On average, people managed about 5 reps at 90% of their max and about 15 at 70%. People varied more at lighter loads. They also got more reps on the leg press than on the bench press.^[11] Moderate evidence.

So a chart that says "80% = 8 reps" is a rough guide, not a rule. A rep range lets you pick the weight from what you can do today. Coaching practice.

For a beginner, almost any sensible program works

ACSM says people new to training benefit from many different programs, as long as they progress and keep at it.^[1] In the 2023 meta-analysis, 101 of 111 comparisons between programs couldn't tell the two apart.^[2] Strong evidence that lifting beats not lifting. The exact rep range matters less than showing up and getting a little better.

How to apply it

The steps below are coaching practice, not a tested finding. They turn the research above into a plan.

1. Think in rep ranges, not percentages. You don't know your 1RM yet, and the reps at any percentage vary from person to person.
2. For muscle, use roughly 5 to 30 reps per set, as long as the set is hard. For most exercises, 6–12 or 8–15 is a comfortable middle. 8–12 is a fine default. It isn't special: ACSM's 2026 stand sets no rep range for muscle.^[1]
3. Let small exercises go higher. Lateral raises and curls work well at 12–20 reps.
4. Keep every set close to failure. Aim for about 1–3 reps in reserve (RIR): how many more reps you could have done before you had to stop. The ACSM panel proposes 2–3.^[1] Going close to failure matters most when the weight is light.^[4]
5. If you want a bigger squat, bench or deadlift, add some heavier sets of about 3–6 reps on that lift. Do this once the movement looks the same on every rep (see learning the lifts).
6. Let the reps choose the weight. If every set hits the top of the range, add a little weight next time. If you can't reach the bottom, take some off. That's progressive overload.
7. Skip the feather-light sets. If a weight lets you do far more than 30 reps, it's probably too light.

A worked example

Here's a full-body session for a 90 kg lifter who wants more muscle and a heavier squat.

• Squat, 3 sets of 4–6: 80 kg × 6, 6, 5, with about 2 reps left in each.
• Dumbbell bench press, 3 sets of 8–12: 24 kg × 12, 10, 9.
• Seated cable row, 3 sets of 10–15: 50 kg × 14, 12, 11.
• Lateral raise, 2 sets of 15–20: 6 kg × 20, 17.

Next session, the squat goes 80 kg × 6, 6, 6. Every set hit the top, so the session after that starts at 82.5 kg.

Now say the gym is busy and the 24 kg dumbbells are taken. Only a 16 kg pair is free. Do sets of 20–25 instead, stopping a rep or two short of failure. That still counts, as long as the sets are hard.

What's practical versus optimal. Light sets work for muscle, but they take longer and feel worse. Heavy sets feel more like a test than a workout. Moderate loads are the practical default, with heavy sets saved for the lifts where you want a bigger max.

Common mistakes

Treating 8–12 as the only range that builds muscle. Many programs use it, so it starts to look like a rule. It's a convenient default. Sets of 6 or sets of 20 build muscle too, if they're hard.

Doing light sets the easy way. Light weights feel safe, and stopping at 15 feels like a full set. But a set of 15 with a weight you could lift 30 times is a warm-up. When you go light, go close to failure.

Going heavy before you can control the lift. Heavy sets look like the serious option. But a weight you can't control is hard to learn with, and growth was similar with lighter loads anyway. Learn the movement first, then add the heavy work.

Testing your max to see if you're improving. It's satisfying to have one big number. You don't need a max test to see progress. Watch your rep records instead: 60 kg × 10 that used to be 60 kg × 8 is progress. Coaching practice.

How Gymbers fits

Programs in the Gymbers app set a target rep range for each set (for example 8–12) and a target RIR for each working set. The weight is yours to choose. The app estimates your one-rep max from logged sets, using the Epley formula. So you can follow your strength without testing a true max.