RightRep EN · PT Open the app
BLOG

Machines vs Free Weights: Which Builds More Muscle? What the Research Says

By RightRep AI Assistant · · 10 min read

Every gym is split down the middle by a line nobody drew. On one side, the free-weight corner: chalk, barbells, and the quiet conviction that machines are furniture for people who aren't serious. On the other, the machine rows: pinned stacks, adjusted seats, and the equally quiet conviction that the squat rack is a lawsuit waiting to happen. Both sides will quote "the science" at you. Neither can say what the machine actually changes about the stimulus.

That is the honest question underneath the tribalism, and it has a cleaner answer than either side expects: for building muscle, the pooled trials say the modality itself barely matters. For building strength, it matters in a specific, testable way. The real difference between a machine and a barbell is smaller than the difference between a hard set and a coasted one — and much smaller than the difference between the program you actually do and the one you abandon.

What the machine actually changes

Three things, mechanically. A machine fixes the path of the resistance: the bar can't drift forward, the handles can't twist, the wobble is engineered out. A barbell or dumbbell makes you steer the load while you move it. And several machines shape the resistance through the range — a cam or pulley that matches the muscle's strength curve, heavier where you're stronger and lighter where you're weaker [1].

That steering is not free. When Saeterbakken and colleagues tested twelve trained lifters on three chest presses with different stability demands, the dumbbell load came out 14% lighter than the Smith machine and 17% lighter than the barbell for the same one-rep max performance — the balance requirement itself eats load [2]. The barbell beat the Smith machine by about 3% [2]. And the extra muscle work that instability demands showed up where you'd expect: not in the target muscles — pectoralis and anterior deltoid activity was statistically identical across all three — but in the stabilizers, with biceps activity climbing neatly from Smith machine to barbell to dumbbells [2]. McCaw and Friday's classic free-weight-versus-machine bench comparison found the same shape: more total muscle activity in the free-weight press, especially at moderate loads [3].

So the machine trades away balance work and buys you a fixed groove. The question is what that trade is worth — and here the pooled literature is unusually consistent.

For muscle size, the answer is a tie

Two independent meta-analyses, published a year apart, pooled the head-to-head training trials. Neither found a hypertrophy difference.

Haugen and colleagues aggregated 13 studies with 1,016 participants, six to twelve weeks of matched training: the effect of free-weight versus machine training on muscle growth was SMD −0.055 (P = 0.751) — statistically indistinguishable from zero, with both modalities growing their groups by similar margins [4]. Heidel's team reached the same wall from the other direction: hypertrophy effect size −0.01, with a confidence interval centered on nothing [5].

The direct trial inside the Schwanbeck study makes the tie concrete. Forty-six adults — 26 of them women, a population the barbell-tribe folklore usually ignores — trained eight weeks on either free weights or machines, every muscle group two to three times a week, three to four sets of four to ten reps. Biceps and quadriceps thickness grew in both groups (P < 0.01) with no difference between them [1]. When volume and effort are equated, the guided path and the bar path produce the same muscle.

For strength, the answer is "which test?"

Strength is where the specificity principle — you get better at what you practice — shows up in the pooled numbers, and it cuts both ways.

In Heidel's meta-analysis, when strength was tested with a free-weight lift, the free-weight groups won (ES 0.655); when tested on a machine, the machine groups won (ES −0.784); and when tested on a neutral device neither group trained on, the two trained identically (ES 0.128, not significant) [5]. Haugen's found the same asymmetry: free-weight tests favored free-weight training (SMD −0.210, P = 0.023), machine tests leaned toward machine training, and the direct like-for-like comparison showed nothing (SMD 0.084, P = 0.387) [4]. You do not get "generally stronger" in the abstract; you get stronger in the groove you trained.

Two wrinkles inside that pattern are worth knowing. First, Haugen's upper-body sub-analysis actually favored machines (SMD 0.245, P = 0.003) [4] — the fixed path may let the target muscle take more of the load instead of spending it on steering. Second, carryover to athletic output was a wash: countermovement jump gains were no different between modalities [4]. If you are judged by a barbell — powerlifting, a max testing day, the big lifts themselves — train the barbell. If you are judged by muscle and general strength, the scoreboard doesn't care.

Why the tie holds: modality is packaging

The reason the hypertrophy result keeps coming back zero is not mysterious. Every trial in those meta-analyses matched the things the dose-response literature says actually drive growth: hard sets, weekly volume, proximity to effort. Volume is the variable with the dose-response curve — every hard set buys growth, each buys slightly less [6][7] — and a hard machine set and a hard barbell set deliver the same payload to the muscle. The modality is the wrapping paper.

This is the same lesson the programming-variables family keeps teaching: frequency stopped predicting growth once volume was accounted for [6], and effort is effort whether the resistance is guided or not. What a machine changes is who can reach a hard set — no spotter needed, no technique barrier, load through the strongest part of the range [1] — not what a hard set is worth.

Five practical rules

  1. Pick the modality you'll keep training. Heidel's team said it outright: for general strength and muscle, choose whichever you prefer and will adhere to [5] — adherence beats every effect size in this article.
  2. Judge the program by payload, not equipment. Hard, direct sets per muscle per week drive growth [6][7], and progression is the same climb either way — a set here, a pin or a plate there [8]. A machine set counts exactly as much as a barbell set, if it's hard and the reps are real.
  3. Train the test. Strength gains are specific to the modality [4][5]. If you compete or test on a barbell, the barbell lifts are the core; machines are accessories, not replacements.
  4. Spend the balance tax deliberately. Dumbbells cost you 14–17% of load compared to guided presses [2]. That's the price of stability work — pay it for that purpose, don't pay it by accident while chasing numbers.
  5. Blend on purpose. Free weights early for the skill and the specificity, machines late to take a muscle to high effort without a spotter [4][5] — the combination covers both columns of the scorecard.

How RightRep fits into this

The camera doesn't know — or care — where the resistance comes from. A rep is graded by the exercise's position profile: joint angles, depth, symmetry, tempo, range of motion. That audit is worth the same on a guided path, which is exactly where it's easiest to fake — the machine fixes the groove, but half-repped partial reps and collapsed tempo still don't count:

  • Every rep gets a 0–100 rep score, machine or barbell. The 25 built-in profiles include the machine-room staples — leg press, lat pulldown, hip thrust — scored by the same on-device engine as the squat and bench.
  • Machine Scan covers the rest of the gym. Photograph any machine the list doesn't have and the AI identifies it and builds a profile for it — scanned machines are scored exactly like the built-in exercises, so "the machine at my gym" is never an excuse for an unaudited set.
  • Form checks keep machine sets honest too. Fixed path, but a real form fault still pulls the score with a stated reason — the seat setting and the half-rep both show up.
  • Volume, load and rest in one dashboard — see how the rep audit stacks up against a classic workout generator on RightRep vs Fitbod or a set logger on RightRep vs Hevy. The full feature list is on the features page.

The bottom line

The research says: for muscle, machines and free weights are a tie [1][4][5]; for strength, you get better at what you practice [4][5]; and the choice that actually moves your results is the one you'll still be making on a Tuesday in week twelve [5]. Choose by preference, train the test if you're tested, and audit the reps either way. This guide pairs with our progressive overload guide (how to progress either modality), how many sets per week (the payload that matters), the rep tempo guide and range of motion vs partial reps (what makes each rep count), rest between sets, the creatine guide, zone 2 training and strength training for longevity. Whichever side of the gym you pick, open the web app, prop up your phone, and let the rep score settle the argument.

This post was written by the RightRep AI Assistant — the same AI that counts and scores your reps at rightrep.app.

Sources

  1. Schwanbeck SR, Cornish SM, Barss T, Chilibeck PD. Effects of Training With Free Weights Versus Machines on Muscle Mass, Strength, Free Testosterone, and Free Cortisol Levels. Journal of Strength and Conditioning Research, 2020;34(7):1851–1859. pubmed.ncbi.nlm.nih.gov
  2. Saeterbakken AH, van den Tillaar R, Fimland MS. A comparison of muscle activity and 1-RM strength of three chest-press exercises with different stability requirements. Journal of Sports Sciences, 2011;29(5):533–538. pubmed.ncbi.nlm.nih.gov
  3. McCaw ST, Friday JJ. A Comparison of Muscle Activity Between a Free Weight and Machine Bench Press. Journal of Strength and Conditioning Research, 1994;8(4):259–264. ovid.com
  4. Haugen ME, Vårvik FT, Larsen S, Haugen AS, van den Tillaar R, Bjørnsen T. Effect of free-weight vs. machine-based strength training on maximal strength, hypertrophy and jump performance – a systematic review and meta-analysis. BMC Sports Science, Medicine and Rehabilitation, 2023;15:103. pubmed.ncbi.nlm.nih.gov
  5. Heidel KA, Novak ZJ, Dankel SJ. Machines and free weight exercises: a systematic review and meta-analysis comparing changes in muscle size, strength, and power. The Journal of Sports Medicine and Physical Fitness, 2022;62(8):1061–1070. pubmed.ncbi.nlm.nih.gov
  6. Pelland JC, Remmert JF, Robinson ZP, Hinson SR, Zourdos MC. The Resistance Training Dose Response: Meta-Regressions Exploring the Effects of Weekly Volume and Frequency on Muscle Hypertrophy and Strength Gains. Sports Medicine, 2026;56(2):481–505. pubmed.ncbi.nlm.nih.gov
  7. Schoenfeld BJ, Ogborn D, Krieger JW. Dose-response relationship between weekly resistance training volume and increases in muscle mass: a systematic review and meta-analysis. Journal of Sports Sciences, 2017;35(11):1073–1082. pubmed.ncbi.nlm.nih.gov
  8. American College of Sports Medicine. Progression models in resistance training for healthy adults. Medicine & Science in Sports & Exercise, 2009;41(3):687–708. pubmed.ncbi.nlm.nih.gov
  9. RightRep. Position profile — how every rep gets scored. RightRep Glossary. rightrep.app