One set per exercise is enough to get stronger when you are new, but two to three sets per exercise produced clearly more in the trials. In a 2026 trial in untrained men, estimated bench press rose 8.5% with one set per exercise and 19.8% with three5.
So the real answer is a range. One set is a working minimum; more sets bought more, with returns that shrink as volume climbs.
What was studied
A set is a run of reps done without stopping. Volume usually means how many hard sets a muscle gets per week. A beginner, in these trials, is someone with no structured resistance training in the months before.
The newest direct test ran in Turkey. Thirty-four sedentary men, average age about 21, were randomised to one set per exercise, three sets per exercise, or no training5. Both training groups did the same seven exercises, twice a week for eight weeks, at 60–70% of estimated 1RM (the heaviest load you could lift once) for 10–12 reps. A trainer supervised every session. By our arithmetic from the protocol, that is 14 sets a week for the one-set group and 42 for the three-set group.
| What | One-set group | Three-set group |
|---|---|---|
| People | 12 | 11 |
| Sets per exercise | 1 | 3 |
| Sets per week (all exercises) | 14 | 42 |
| Sessions | twice a week, 8 weeks | twice a week, 8 weeks |
| Load | 60–70% of 1RM, 10–12 reps | same |
Strength was estimated, not tested with a true maximum lift: the men lifted loads until six reps were no longer possible, and a formula converted that to a 1RM5. Size was measured with a tape (arm and thigh girth) and a bioimpedance scale for lean mass, not with a scan.
In plain terms: same exercises, same loads, same coach; the only planned difference was one set or three.
What the studies found
Strength. In the 2026 trial, estimated bench press rose 8.5% with one set and 19.8% with three; squat rose 3.6% and 14.8%5. The control group's estimates dipped slightly. For the squat, three sets beat one set significantly; for the bench, three beat one and both beat control.
Change in estimated 1RM over 8 weeks, untrained men · %
Size. Upper-arm girth rose 0.41 cm with one set and 1.71 cm with three, against 0.09 cm in the control group5. Lean mass rose in both training groups, more with three sets. Thigh girth did not differ between groups.
Doing something beats doing nothing. On every strength, power and fitness test in the trial, the one-set group improved more than the controls5. The authors wrote that "in novice individuals, even a single set may provide a sufficient stimulus to induce meaningful adaptations, whereas higher volumes may be required to elicit further improvements."
Longer and larger evidence points the same way. A 2015 trial followed 48 men with no lifting experience for six months, three sessions a week, in groups doing 1, 3 or 5 sets per exercise2. Elbow-flexor thickness increased significantly only in the 3- and 5-set groups, with 5 sets ahead. All training groups gained lean mass, with no difference between them.
A 2010 meta-analysis pooled 8 studies comparing one set with several1. Multiple sets were linked to about 40% larger hypertrophy effect sizes than one set, in both trained and untrained people. The effect sizes rose step by step: 0.24 for one set, 0.34 for 2–3 sets and 0.44 for 4–6 sets per exercise.
Muscle growth effect size by sets per exercise · effect size
In plain terms: one set works, and every direct comparison found more sets did more, at least up to a few per exercise.
More, but with diminishing returns
How far does "more" go? A 2017 meta-analysis of 15 studies found each extra weekly set was associated with about 0.37% more growth3. The largest analysis so far, from 2026, pooled 67 studies and 2,058 participants4. Growth and strength both rose with weekly volume, but with diminishing returns: each added set bought less than the one before. The flattening was stronger for strength than for size.
For a beginner, that curve has a practical reading. The first sets of the week carry the biggest share of the gain. The step from one set to three, which is where the 2026 trial worked, sits on the steep part of the curve. Going far beyond that has less evidence of extra payoff, and none of these beginner trials tested very high volumes; for trained lifters, see what happens if you double your training volume. For the weekly numbers researchers use, see how many sets per muscle group per week.
In plain terms: the first few sets do most of the work. More still helps, but each extra one helps less.
Why one set might be the right start anyway
The trials measured muscle and strength, not whether people kept training. A one-set session is short: the 2026 one-set group did seven sets per session5. The authors saw a place for it: single-set training "may remain sufficient for time-efficient health-oriented exercise", while multi-set programmes "appear more effective for maximizing performance-related and structural adaptations". Whether a shorter start helps people stick with training was not tested here. The authors also note their result reflects total training dose, not set count alone, since three sets also meant three times the reps.
A typical beginner plan builds volume over weeks rather than starting at the top (see beginner workout program and what progressive overload is).
The studies at a glance
| Study | Who was studied | Key result | Year |
|---|---|---|---|
| Krieger, *J Strength Cond Res*1 | 8 studies, trained and untrained | Multiple sets gave about 40% larger growth effects than one set | 2010 |
| Radaelli et al., *J Strength Cond Res*2 | 48 untrained men, 6 months | Arm thickness rose only with 3 and 5 sets; 5 sets gained most | 2015 |
| Schoenfeld et al., *J Sports Sci*3 | 15 studies, weekly sets vs growth | Each extra weekly set linked to about 0.37% more growth | 2017 |
| Pelland et al., *Sports Med*4 | 67 studies, 2,058 participants | Growth and strength rise with volume, with diminishing returns | 2026 |
| Coskun and Kafkas5 | 34 untrained men, 8 weeks | Bench +8.5% with one set vs +19.8% with three; one set beat no training | 2026 |
Limits worth knowing
- Small and short. The 2026 trial had 11 or 12 men per group for eight weeks5.
- Men only. Both beginner trials studied young men; women and older adults were not included.
- Estimated strength, tape-measured size. No true 1RM and no muscle scan in the 2026 trial. Its abstract mentions a cross-sectional-area result that its methods and tables do not support, so we left it out.
- Unedited manuscript. We read the 2026 paper in full, as an accepted manuscript before copy-editing; it has small internal inconsistencies. It was registered after the trial ran, and it reported no funding.
- Abstract only. We read the 2010, 2015, 2017 and 2026 meta-analyses and trials other than Coskun as abstracts.
- Effort not matched. One set and three sets may differ in how hard each set was; the authors list this as a limit.
Latest research
Two 2026 papers frame the question. The Coskun and Kafkas trial, published in July 2026, is the newest direct test of one set against three in people starting from zero5. The Pelland meta-regressions, the largest pooled analysis of volume, describe the shape of the curve: up, then flattening4. Neither tells a beginner exactly where their own curve bends.
The short version
- One set per exercise made untrained men stronger than no training at all.
- Three sets per exercise produced clearly more strength and arm growth than one in the same eight weeks.
- Across studies, more sets meant more growth, with each extra set adding less.
- The trials were short and in young men; very high volumes were not tested in beginners.
Educational overview only — not medical advice.
