Training volume for hypertrophy: sets per muscle per week

An evidence-grade synthesis of the dose-response relationship between weekly resistance-training volume and muscle hypertrophy, with attention to effect sizes, individual variability, and the upper-bound question.

Background

Skeletal muscle hypertrophy in response to resistance training is a function of mechanical tension applied to motor units recruited during a working set, accumulated across sufficient sets per session and sessions per week to exceed the recovery capacity of the trained muscle. Of the manipulable training variables — load, proximity to failure, frequency, exercise selection, and volume — weekly volume has emerged in the meta-analytic literature as the variable with the most robust dose-response signal for hypertrophy. This article summarizes what the trial-level and meta-analytic data actually support.

The Schoenfeld 2017 meta-regression

The most cited single source on this question is the Schoenfeld, Ogborn, and Krieger (2017) meta-analysis published in the Journal of Sports Sciences. The authors aggregated 15 controlled trials in which resistance-training volume was manipulated as the independent variable and muscle hypertrophy was an outcome (typically measured by ultrasound, MRI, or DXA). The meta-regression detected a graded relationship between weekly working sets per muscle group and percentage increase in muscle thickness. Studies in which subjects performed fewer than 5 weekly sets per muscle showed an average effect size around 5.4 percent muscle thickness gain over the trial duration (typically 6 to 12 weeks), studies at 5 to 9 weekly sets averaged around 6.6 percent, and studies at 10 or more weekly sets averaged around 9.8 percent.

The interpretive caveats matter. The meta-regression aggregates across heterogeneous training programs, training experience levels, measurement instruments, and trial durations, so the absolute effect sizes are imprecise. The 10+ sets category is also the bin in which the most data accumulated, which improves precision at the high end and reduces it at the low end. The trend across the three bins is robust, but the precise shape of the dose-response within each bin is not.

Higher volume in trained men: the 2019 follow-up

Schoenfeld and colleagues (2019) tested the upper-end question more directly in a controlled trial of 34 resistance-trained men randomly assigned to one of three weekly volume groups (low: 1 set per exercise; moderate: 3 sets per exercise; high: 5 sets per exercise) for 8 weeks across a six-exercise lower-body and upper-body program. Higher volume groups produced greater muscle thickness gains, with the largest effects in the high-volume group. The trial supports the position that the dose-response continues into the higher-volume range in trained populations — at least within the 8-week trial window.

A separate trial by Heaselgrave et al. (2019), in 49 trained men randomized across three weekly elbow flexor volume conditions (9, 18, or 27 sets per week), found broadly similar gains in elbow flexor thickness across the 9 and 18 set conditions and a less favorable result at 27 sets, with elevated soreness and reduced perceived recovery in the highest-volume group. The systematic review by Baz-Valle et al. (2022) reached a similar conclusion across the broader trial literature: weekly volumes in the 10 to 20 working-set range cover the responsive range for most trained individuals, and very high volumes (above 20 to 25 sets per muscle per week) produce diminishing or reversing returns in some trials.

The volume-landmarks framework

A practical operationalization commonly used by hypertrophy-focused coaches frames weekly volume in terms of three landmarks:

• MEV (minimum effective volume): the lowest weekly set volume per muscle that produces measurable hypertrophy. For most trained individuals this falls in the 4 to 8 weekly set range.
• MAV (maximum adaptive volume): the volume range in which hypertrophy is maximized given recovery. This falls in the 10 to 20 weekly set range for most muscles in trained individuals, with substantial individual variability.
• MRV (maximum recoverable volume): the upper bound beyond which additional volume produces accumulated fatigue greater than the recovery capacity of the trainee. This is highly individual and is rarely useful as a fixed prescription.

The MEV-MAV-MRV framework is consistent with the meta-analytic dose-response data and offers a useful structure for prescription. It is not an independently validated set of physiological constants; the landmarks are heuristics derived from the same trial literature summarized above and from coaching practice.

Effect sizes and individual variability

The meta-analytic effect sizes — single-digit percentage muscle-thickness changes over 6 to 12 week trials — are the average across substantial individual variability. Trial-level data consistently show wide between-subject variance in hypertrophy response, with some subjects gaining little to no muscle thickness on a protocol that produces double-digit gains in others. The sources of this variability include genetic factors, baseline training status, dietary protein adequacy, sleep, age, sex, and the precision with which the prescribed program was actually executed. Individual response variability is large enough that meta-analytic central tendencies should be interpreted as starting points for prescription, not as expected outcomes for any specific trainee.

Interaction with effort and proximity to failure

Volume cannot be considered independently of the effort applied to each set. A trial that prescribes 20 weekly working sets at 6 reps in reserve is not comparable in stimulus to a trial that prescribes 10 sets at 0 to 1 reps in reserve, even though the set count differs by a factor of two. The Helms et al. (2023) work on RPE-based loading underscores the broader point that load prescription, proximity to failure, and volume are measured in different units and combined in non-trivial ways at the level of the actual training stimulus. Volume recommendations in this article assume working sets taken to within roughly 0 to 4 reps in reserve, which is the operating range used in most of the cited trials.

What the evidence does not support

The literature does not support a single optimal weekly volume that applies across populations and across muscles. It does not support extreme high-volume protocols (above approximately 25 sets per muscle per week) as superior on the available trial evidence. It does not address very long timeframes (training years rather than training weeks) at which different volumes may produce different cumulative outcomes. The strongest defensible position is that 10 to 20 weekly working sets per muscle covers the empirically responsive range for most trained individuals, with adjustment toward the lower end for novices and individuals with limited recovery capacity, and toward the higher end for advanced trainees willing to manage the recovery cost.