REF-1785 · Sato et al., 2026 · Journal of Strength and Conditioning Research, · Paper · PubMed

Single-leg and two-leg leg press gave similar gains, with no significant group difference, in two-leg leg-press 1RM (+14.7% vs +13.4%), knee-extensor torque and squat, countermovement and drop jumps over 8 weeks, while single-leg training raised single-leg 1RM more (+22.7% vs +13.7%, p < 0.01), in a trial of 28 resistance-untrained young men (14 per group) using eccentric-accentuated sets

Main caveat: One small trial, 14 per group, resistance-untrained young men only; no non-training control.
moderate evidenceMeasured on non-athletes and recreational peopleRead from the abstract only

Sato and colleagues compared eight weeks of leg-press training done with one leg at a time against the same training done with both legs together. Twenty-eight resistance-untrained young men were split into two groups of 14 and trained twice a week, 3 sets of 10 reps with a 3-s lowering and a 1-s lift, the load progressing from 30% to 120% of pretraining 1RM.

Two-leg leg-press 1RM rose similarly in both groups (two-leg group 150.4 to 170.0 kg, +13.4%; single-leg group 144.3 to 165.4 kg, +14.7%), as did knee-extensor isometric, concentric and eccentric torque and squat, countermovement and drop jump height, with no significant group differences. Single-leg leg-press 1RM rose more in the single-leg group (+22.7% vs +13.7%, p < 0.01).

Changes in strength did not correlate significantly with changes in jump height. The authors attribute the single-leg 1RM difference to training specificity and conclude that nothing else changed differently between the two methods.

Participants

28 resistance-untrained young men; 14 bilateral, 14 unilateral.
Age, body size, other sport or activity, and dropouts: not in the abstract.

Protocol

Both groups: leg press, 3 sets of 10 eccentric-accentuated contractions (3-s eccentric, 1-s concentric) per session, twice a week for 8 weeks; intensity raised from 30% to 120% of pretraining 1RM (which 1RM, two-leg or single-leg, served as the reference for the single-leg group is not in the abstract).
Unilateral group: each leg trained separately within the session; bilateral group: both legs together.
Tests before and after: bilateral and unilateral leg-press 1RM; maximal voluntary contraction (isometric, concentric, eccentric) torque of the knee extensors and flexors; squat, countermovement and drop jump height.
Not in the abstract: randomisation method, blinding, rest between sets, whether a 120% load was possible because of the eccentric emphasis or the equipment, knee-flexor results, effect sizes and confidence intervals for the group differences.

Key Results

Outcome, change over 8 weeksBilateral group (n = 14)Unilateral group (n = 14)Group difference
Bilateral leg-press 1RM150.4 ± 18.7 → 170.0 ± 21.1 kg, +13.4 ± 9.3%144.3 ± 17.8 → 165.4 ± 21.2 kg, +14.7 ± 7.1%not significant
Unilateral leg-press 1RM+13.7 ± 8.3%+22.7 ± 7.0%p < 0.01
Knee-extensor MVC, isometric+10.2 ± 7.6%+12.4 ± 9.3%not significant
Knee-extensor MVC, concentric+11.5 ± 12.0%+7.0 ± 8.1%not significant
Knee-extensor MVC, eccentric+10.7 ± 10.8%+11.0 ± 7.5%not significant
Squat jump height+14.0 ± 13.4%+9.8 ± 13.0%not significant
Countermovement jump height+9.6 ± 11.7%+9.0 ± 8.7%not significant
Drop jump height+14.2 ± 16.4%+16.3 ± 19.4%not significant
All within-group changes listed were significant (p < 0.05).
No significant correlations (p > 0.05) between changes in strength and changes in jump height.

Practical Application

A single-leg leg press built two-leg strength and jump height with no significant difference from the two-leg version in beginners, and built more single-leg strength (+22.7% vs +13.7%). One trial of 14 per group in resistance-untrained young men; "no significant difference" in a trial this size does not rule out a modest difference, and the abstract gives no intervals.
Bears on [SYNTH-107]'s home rule that single-leg variants are a way to load the legs when weights are light; that guide cites this trial at its home-weights default and keeps the single-leg rule as practice. Inference, not tested here: a single-leg exercise needs less external load per set than its two-leg version, which is why it suits light home weights; the loads lifted are not in the abstract, and this trial used a machine leg press, not dumbbell split squats or step-ups.
What it does not say: whether single-leg work helps cycling or running. SYNTH-43 keeps heavy two-leg lifts as the default for steady events on REF-148, a pilot in which two-leg lifting improved economy where alternating single-leg lifting did not; this trial measured no endurance outcome and does not change that.
Match the test to the goal (inference): the larger single-leg gain was specific to the single-leg test, so judge progress on the movement trained.

Limitations

One small trial, 14 per group, resistance-untrained young men only; no non-training control.
The abstract says the men were "assigned"; randomisation comes from the indexing, and its method, blinding of testers and dropouts are unknown.
The large standard deviations (for example jump changes of ±13-19%) mean between-group differences of a few percent would not be detected.
Eccentric-accentuated leg press only; how far this transfers to faster or free-weight single-leg lifts is unknown.
Written from the abstract; full text not read.
How this entry was checked and changed
Why this level

a parallel trial with 14 per arm (the rubric's 10-19 per arm band), outcomes are strength and jump performance; allocation method, blinding and dropouts are not in the abstract