Free Weights vs Cables vs Machines: What Actually Builds Muscle

Are machines as good as free weights? For building muscle size, the research says they are close enough that equipment is rarely your bottleneck. A 2023 meta-analysis found no detectable difference in hypertrophy between the two, and the real differences show up in strength specificity, safety near failure, and practical convenience.

That is the short answer to the free weights vs machines vs cables hypertrophy debate. The rest of this article covers what the studies actually measured, where cables fit, and how to mix all three in a split without turning it into a religion. (This sits inside our wider exercise selection and biomechanics guide, which is worth reading if you are building a program from scratch.)

Barbell purism is an old gym habit. The data does not back it up.

Free weights vs machines vs cables: what the best evidence says

The biggest piece of evidence is Haugen et al. (2023), a systematic review and meta-analysis of studies that directly compared free-weight and machine training for at least 6 weeks in adults aged 18-60. It pooled 13 studies and 1016 participants (789 men, 219 women). Only 5 of those looked at hypertrophy, so keep that in mind.

Here is what it found.

• Hypertrophy: no difference between modalities (SMD -0.055, 95% CI -0.397 to 0.287).
• Dynamic and isometric strength in a direct comparison: no difference.
• Jump performance: no difference.
• Strength specificity: strength on free-weight tests rose a little more after free-weight training (SMD -0.210), and strength on machine tests tended to rise more after machine training (SMD 0.291, not quite significant).

The authors concluded that strength changes are specific to the training modality, and that the choice between free weights and machines comes down to individual preferences and goals.

Why the hypertrophy result is believable

It is not one lucky paper. Several smaller trials point the same way, each with its own limits.

• Schwanbeck et al. (2020) randomised 46 people (26 women, average age 22) to free weights or machines for 8 weeks. Biceps and quadriceps thickness rose with no difference between groups. Machine bench press strength rose more in the machine group (13.9 vs 8.6%), while free-weight bench, squat and Smith machine squat strength rose similarly in both.
• Hernandez-Belmonte et al. (2023) took 34 resistance-trained men and trained them for 8 weeks on a barbell or machine version of squat, bench, bench pull and shoulder press, with intensity matched using bar velocity. No between-group differences in muscle architecture or athletic measures. The free-weight group improved change of direction and some balance tests, and the machine group meaningfully improved upper-limb anaerobic power.
• Amanuma et al. (2025) trained each leg of 8 untrained young women differently for 9 weeks, lunges on one side and incline leg press on the other. Ultrasound showed comparable thickness gains in the vastus lateralis and rectus femoris. The sample was tiny and untrained, so treat it as supporting evidence, not proof.

One caveat that matters. Most of these trials ran 8-10 weeks in young adults. Over years of training, other factors may creep in, and nobody has run a ten-year free-weight versus machine trial.

Why would machines work as well as barbells?

Because muscle does not know what equipment it is pulling against. It responds to mechanical tension, and tension can be produced by a barbell, a cable stack or a plate-loaded sled.

Two things seem to matter far more than the tool.

First, how hard the set is. In Refalo et al. (2023), a meta-analysis of 15 studies, there was a trivial advantage for training to any definition of set failure, but no evidence that training to momentary muscular failure beat stopping short. In plain terms: sets should be hard, but you do not need to grind every rep into the ground.

Second, load range. A review by Schoenfeld and colleagues (2021) argues that muscle adaptations can be obtained across a wide spectrum of loading zones, not only a narrow “hypertrophy range”. That is one reason light-to-moderate loads on a cable work.

So when you compare tools, compare them at matched effort. A sloppy dumbbell set stopped 6 reps short loses to a hard machine set, and the reverse is also true. Our RIR and RPE calculator helps you judge how close to failure a set really was.

Free weights vs cables: do cables build muscle as well?

Cables have the thinnest direct evidence of the three. There is no large meta-analysis of cables against dumbbells, so be careful with anyone who says otherwise.

The closest controlled data I would point to is Nunes et al. (2020). Thirty-five young adults trained the biceps for 10 weeks, three times a week, with preacher curls for 3 sets of 8-12. One group used a cable, where torque is higher when the elbow is flexed. The other used a barbell, where torque is higher when the elbow is extended and the biceps is stretched.

• Biceps thickness rose 7% on cable and 8% on barbell, not a meaningful difference.
• Strength gains were greater for the barbell only at the longest muscle length (39% vs 30% at 20 degrees of elbow flexion).

That tells you something useful. The bar and the cable load the muscle differently, and the growth was still similar in this group. It does not tell you the same would hold for the chest, back or shoulders, and it does not tell you what happens in advanced lifters.

Where cables earn their place

These are practical reasons, not proven growth advantages.

• Constant direction of pull. On a dumbbell lateral raise, the load is easiest at the bottom. A cable can keep tension on the side delts at the start of the rep. We cover pulley setup in our cable setup guide.
• Small jumps, sometimes. Stack plates are often 5 kg or more, but many stacks take add-on weights or magnetic plates, so you can progress in smaller steps.
• Easy to train one side at a time. Single-arm cable work is simple to set up.
• Safe to take hard. There is no bar to drop.

The downside is busy-gym reality. In many commercial gyms the cable station is the most crowded spot, so the “best” cable setup is the one with a free pulley at 6 pm. That is a good reason to have a dumbbell fallback.

Does stability or instability change muscle growth?

This is the main argument for free weights: the extra balance demand recruits stabilisers.

Behm and colleagues (2006) reviewed the literature on instability in resistance training. They note that instability can decrease the externally measured force output while maintaining high muscle activation, and that the lower force may be detrimental to absolute strength gains. Some studies also report increased co-contractions with unstable training. For health and rehabilitation programs, they suggest including both stable and unstable exercises.

Two cautions. This is a review, not a hypertrophy trial, and it deals with unstable surfaces and devices more than with barbell versus machine. Second, more muscle activity in stabilisers is not the same thing as more growth in the muscle you want to train.

A practical way to hold this: if the stabiliser is not the muscle you are training, it can become the weak link that ends the set early. This is the logic behind our article on external stability in exercise design.

The practical advantages that actually matter

If growth is similar, the choice becomes about everything else.

Factor	Free weights	Cables	Machines
Size (hypertrophy)	Comparable	Comparable (limited direct evidence)	Comparable
Strength on that specific lift	Best carryover	Limited data	Best on the machine itself
Skill demand	Higher	Moderate	Lower
Safety near failure	Needs spotter or safeties	Generally good	Generally good
Load jumps	Often 2.5 kg plates	Depends on stack	Depends on stack
Setup	Easy	Fiddly	Fit varies by body

Safety near failure

Free weights ask you to bail out of a failed rep. A heavy barbell squat or bench to failure alone is a bad plan, which is why people stop two or three reps short. A machine lets you chase effort without that overhead.

I am not citing a trial that proves lower injury rates on machines, so treat this as common-sense gym logic. If you are tired, training alone, or working near a joint that is acting up, the fixed path is easier on your nerves.

Microloading

Free-weight progress is often limited by plate sizes. Going from 20 kg dumbbells to 22 kg is a 10% jump. Some cables and machines take add-on weights for smaller increments, and fractional plates exist if your gym has them. Our article on progressive overload on machines vs free weights goes into how to track it.

Machine fit

A machine built for an average body can fit you badly. If a seat cannot be adjusted to your height or the pad lands on a joint instead of a muscle, skip it. That is not a hypertrophy loss, that is a bad tool.

How to mix free weights, cables and machines

You do not need a formula. Here is a simple pattern that follows the evidence above.

1. Start with a heavy compound while you are fresh: squat, row, press, or a machine version if that is what you can progress.
2. Add a stable machine or cable movement for the same muscle, taken close to failure.
3. Finish with isolation on a cable or machine, where the target muscle is the limiting factor.

A sample day, written in plain terms.

• Incline dumbbell press, 3 sets of 6-10
• Machine chest press, 2-3 sets of 8-12, near failure
• Cable lateral raise, 3 sets of 12-15
• Overhead cable triceps extension, 2-3 sets of 10-15

On our free workout plans you will see this pattern applied across a full split. If your gym is crowded and you need alternatives, see equipment shortage substitutions.

Budget and home training

If you train at home on a budget, a pair of adjustable dumbbells, a bench and a resistance band cover a lot of what a gym does for muscle growth. Bands are not cables, but they offer a similar direction-of-pull idea. Do not feel pressure to buy a rack and cable tower to build muscle. Evidence does not require it.

Common mistakes in the equipment debate

Other patterns I see often.

• Machines for the ego, free weights for the ego. Either way, you are choosing for identity, not for the target muscle.
• Sloppy reps on cables. Swinging the stack turns a tension exercise into a momentum exercise.
• Avoiding free weights completely while wanting a bigger squat or bench. The strength specificity finding applies.
• Avoiding machines completely and then getting stuck because a joint complains.

Limits of the evidence

It would be dishonest to stop at “parity” without saying where the data is thin.

• Only 5 of the 13 studies in the main meta-analysis measured hypertrophy.
• Most trials ran 6-10 weeks, mostly in young adults, many untrained. Long-term outcomes in advanced lifters are less certain.
• Few trials compare cables directly, so the cable story rests on mechanism, one biceps study, and practical experience.
• “Machine” covers dozens of designs, from plate-loaded presses to cam-based stacks. Results for one may not carry to another. For how cam shape plays into this, see selectorized machine cam profiles.

If you have a joint injury, ongoing pain, or a medical condition, see a doctor or physiotherapist about which loading is appropriate before adding heavy sets on any equipment.

So which should you choose?

Take the tool you can train hard, progress and repeat for months. If you love barbells, use barbells. If your knees prefer a leg press, use the leg press. If a cable gives you a better feel for your side delts, use that.

Use free weights for the lifts you want to be strong at. Use cables and machines to add volume without the fatigue and risk of another heavy compound. Track your sets and weights, and let the numbers, not tradition, decide if it works.

For a related look at fixed-path barbells, read our piece on the Smith machine vs free barbell, and for choosing exercises by effort versus fatigue, see stimulus-to-fatigue ratio.