All articles
TrainingNovember 18, 2026·7 min read·By Ossian

Machines vs Free Weights for Hypertrophy: What Studies Show

In the studies that compared machines and free weights, the two built about the same amount of muscle. Two meta-analyses and three trials found no difference in muscle growth. Strength gains tend to be larger on the kind of equipment you train with.

In the studies that compared machines and free weights, the two built about the same amount of muscle. Two meta-analyses and three trials found no difference in muscle growth. Strength gains tend to be larger on the kind of equipment you train with.

So for size, either kind of equipment works. Pick an exercise for each muscle that you will keep doing and push hard, on a machine or with free weights. A 2026 trial gives one reason to consider machines. After a Smith machine squat session, most measures of neuromuscular fatigue in the thighs were lower than after the same free-weight squat session, with no significant difference in bar speed loss or in how many reps the lifters felt they had left (Marcos-Frutos et al., 2026). If you prefer machines, the choice is wide. Leaving out bodyweight exercises, GainsLog's library has 243 exercises on a machine, cable or Smith machine and 122 with a barbell, dumbbell or kettlebell. For chest and back there are about three guided options for every free-weight one.

Do machines build as much muscle as free weights?

In the studies we read, yes. Each row below compared the two directly and measured muscle size. None found a significant difference, though small studies and wide confidence intervals cannot rule out a modest one.

Studies comparing machines and free weights for muscle growth
StudyWho and how longMuscle growth
Haugen et al. 2023, meta-analysis13 studies, 1,016 adults; 5 of the studies measured muscle sizeNo difference
Heidel et al. 2022, meta-analysisLiterature searched up to September 2020No difference
Hernández-Belmonte et al. 202338 trained men, 8 weeks, barbell or machine versions of 4 liftsNo difference in quads, chest or abs
Schwanbeck et al. 202046 young adults (26 women), 8 weeks; also in Haugen's analysisNo difference in biceps or quads
Amanuma et al. 20258 untrained women, 9 weeks, lunge on one leg, leg press on the otherNo difference in rectus femoris or vastus lateralis

What the trials did

Hernández-Belmonte and colleagues ran the only one of the three trials in people who already lifted. Thirty-eight trained men followed identical programs for 8 weeks, and the two groups differed only in whether they did the full squat, bench press, prone bench pull and shoulder press with a barbell or on a machine. Cross-sectional area of the quads, chest and abs grew in both groups, and the difference between them was 2 percentage points or less and not significant. Joint stiffness, pain and disability scores fell in both groups (Hernández-Belmonte et al., 2023).

Schwanbeck and colleagues trained 46 young adults for 8 weeks, each muscle 2 to 3 times a week for 3 to 4 sets of 4 to 10 reps. Biceps and quad thickness increased with no difference between the free-weight and machine groups (Schwanbeck et al., 2020).

Amanuma and colleagues removed the differences between people by training both of each woman's legs, one with lunges and the other on an inclined leg press, 3 times a week for 9 weeks. The authors report growth on both legs at every site they measured, though one site on the leg press leg just missed their significance threshold. At the lower part of the rectus femoris, muscle thickness rose 24.8% on the lunge leg and 27.3% on the leg press leg, and no site differed between the two (Amanuma et al., 2025).

The two meta-analyses pooled earlier studies of this kind, including Schwanbeck's in Haugen's case, and found no difference in muscle growth between free weights and machines (Haugen et al., 2023; Heidel et al., 2022). The evidence has limits. The three trials above ran 8 to 9 weeks with 8 to 46 people each, only five of the 13 studies in Haugen's analysis measured muscle size, and the pooled estimates leave room for a modest difference either way.

Where free weights may help: strength on free-weight lifts

Strength gains tend to be larger on the equipment you trained with. In Haugen's meta-analysis, strength measured on free-weight tests rose more with free-weight training, and strength on machine tests tended to rise more with machine training, though that second result was not significant (Haugen et al., 2023). Heidel's meta-analysis found the effect in both directions, and similar gains when strength was tested on a neutral device (Heidel et al., 2022).

The single trials are mixed. In Schwanbeck's, machine bench press strength rose 13.9% in the machine group and 8.6% in the free-weight group, while free-weight bench press, squat and Smith machine squat strength rose by about the same in both groups (Schwanbeck et al., 2020). In Hernández-Belmonte's, each group improved most on the version it trained, but across all eight strength tests the two groups gained about the same (Hernández-Belmonte et al., 2023).

If you want a heavier barbell bench press or squat, train those lifts with a barbell, since both meta-analyses found larger gains on free-weight tests from free-weight training. If the goal is how your chest and legs look, the trials give no reason to prefer the barbell.

What about stabilizers and fatigue?

The case for free weights is that an unstable load recruits more muscle, as Schwanbeck and colleagues put it when they set up their trial. Their own trial found no difference in growth between the groups (Schwanbeck et al., 2020). None of these studies tested that argument directly.

A trial published online in October 2026 measured fatigue instead. Marcos-Frutos and colleagues had 19 trained men do 4 sets of 9 squats at 70% of their one-rep max, once as a free-weight squat and once on a Smith machine. Bar speed loss and how many reps the lifters felt they had left did not differ significantly. After the free-weight session, most measures of fatigue in the thigh muscles were higher, both in the muscle itself and in the nervous system's drive to it. Since earlier trials show similar adaptations from the two, the authors suggest the Smith squat may be a more efficient way to build strength with less neuromuscular fatigue (Marcos-Frutos et al., 2026). It was one session per condition, and it measured fatigue, not growth.

How to choose for each muscle

Ossian's note "Good machines are optimal for hypertrophy" says that muscles respond to tension and don't know whether it comes from a machine, a dumbbell or a cable. His note on keeping gym training simple says to pick one exercise per muscle group, preferably a machine or Smith machine, to maximize enjoyment and effectiveness. The trials above found no difference in growth, so they do not show machines ahead.

His log from his summer 2024 cut is mostly machines. He trained full body five times a week with about 2 sets per muscle per session and 1 to 2 reps in reserve, and 20 of the 30 exercises in the split used a machine, cable or Smith machine. The rest were free-weight and bodyweight lifts such as dumbbell presses and pull-ups. He called it the best hypertrophy training he had done while cutting.

In practice, pick an exercise for each muscle that you will keep doing, and log it so you can see whether the weight or reps go up. The guide to RIR and training to failure covers how close to failure to go. The table shows how many options GainsLog's library has for each muscle, and the machine exercises by muscle and dumbbell exercises by muscle pages list them.

Exercises in GainsLog's library for nine of its twelve muscle groups, by primary muscle, classified by the equipment in each name, or the app's own tag where the name gives none (bodyweight exercises and the hip adductor machine, which the app files under quads, left out)
MuscleMachine, cable or SmithBarbell, dumbbell or kettlebell
Chest3912
Back4613
Shoulders2618
Biceps2324
Triceps4317
Quads188
Hamstrings94
Glutes1412
Calves70

When free weights are the better pick

Pick free weights when you want to get stronger on a barbell or dumbbell lift, when your gym has no good machine for a muscle, or when you train at home. Heidel and colleagues advise people training for strength or power to consider how that strength will be tested and used (Heidel et al., 2022). The resistance profile guide covers how to judge one exercise against another, whatever the equipment.

GainsLog's program builder asks "What do you train with?" and offers full gym, home gym, dumbbells only and calisthenics. With home gym it leaves out most machine, cable and Smith machine exercises. You can replace any exercise, and the app suggests swaps that train the same movement, so a barbell lift can become a machine one and back. To follow a program built around the equipment you have and log every set, download GainsLog. The program builder is part of Premium.

How this article was made

Researched and drafted with AI assistance. The preference for machines and the 2024 training split come from Ossian's notes and training log. We read the studies in their published abstracts, counted the exercise library from GainsLog's app code and this site's exercise pages, and checked the sources on 18 November 2026.

Share this articleLet a training partner know what we are building.