ACSM’s New Resistance Training Rules: What Actually Changed

Fifteen years ago, the standard hypertrophy prescription came with a heavy-load bias: roughly 70% of one-repetition maximum (1RM) or more, with increasingly elaborate programming for increasingly experienced lifters. The newest American College of Sports Medicine position stand changes that picture. For muscle growth, loads as low as 30% 1RM can work when effort is high. Training to failure depends on the load. And several gym-floor obsessions—set structure, time under tension, equipment choice, and named periodization schemes—have not shown consistent superiority over simpler progressive training.

That does not mean programming details never matter. It means the evidence does not support treating every detail as a growth switch. The biggest caveat is equally important: much of the evidence comes from inexperienced trainees. Experienced lifters should treat broad claims about advanced programming as provisional, not as permission to throw out everything that has worked for them.

The 70% rule lost its claim to be a minimum

The 2026 ACSM position stand synthesized 137 systematic reviews and meta-analyses in an umbrella review designed to update the 2009 guidance. Its conclusion on hypertrophy is practical: a broad range of loads, from at least 30% to 80% 1RM, can build muscle when sets demand high effort. Heavier loading still has a different job. The stand recommends at least 80% 1RM for maximal strength development.

A 2023 systematic review by Weakley and colleagues supports the lower end of that range: loads below 50% 1RM produced comparable hypertrophy to higher loads when participants trained close to concentric failure. Light loads are therefore a usable option for people with limited equipment, joint irritation, or an injury-management plan. They are also usually more uncomfortable at the end of a set. A set of 30% 1RM does not become easy because the dumbbell looks polite.

The condition matters. These findings do not say load is irrelevant. They say the old 70% threshold was too restrictive as a hypertrophy minimum. A light set stopped far from failure may not provide the same stimulus as a heavier set or a light set pushed close to failure.

Failure is a tool, not a commandment

The cleanest reason to avoid a universal rule about failure comes from a randomized trial by Lasevicius et al. In that study, training to failure promoted greater hypertrophy with low loads of 50% 1RM or less, but did not add hypertrophy under high-load training of at least 70% 1RM. The result fits a practical rule: when using light weights, take sets to or near failure; with heavy weights, stopping about one to three repetitions short can produce comparable muscle growth.

Still, this is one randomized controlled trial, not a population-wide effect estimate. It cannot settle the ideal failure prescription for every exercise, training history, or weekly volume. A separate bench-press trial by Rodiles-Guerrero et al. compared 0%, 15%, 25%, and 50% velocity-loss thresholds and found different strength, neuromuscular, and hypertrophy adaptations. The near-failure 50% threshold did not dominate every outcome. Because the study examined one upper-body exercise and muscle group, it cannot prescribe velocity loss for squats, deadlifts, or an entire training program.

The reasonable conclusion is narrower than either camp prefers. Failure is useful when low loads need more stimulus. It is optional when heavier loads already make the target muscle work hard enough.

Many programming arguments shrink after volume and effort match

The ACSM review found no consistent advantage for particular periodization schemes, time under tension, blood-flow restriction (BFR), equipment type, or set structure over simpler progressive overload. An independent 2024 umbrella review by McLeod et al., examining load, volume, frequency, failure, rest intervals, and set structure in healthy adults, reached a similar broad conclusion: complex prescription variables often produced null or highly variable findings once basic effort and volume were accounted for.

That pattern does not make these variables useless. Cluster sets may help manage fatigue. A machine may be more comfortable than a barbell. Periodization may organize training around a competition or a stressful work season. But the evidence does not show that these choices reliably create extra hypertrophy by themselves.

A 2025 randomized trial of cluster versus traditional sets illustrates the point. In resistance-trained participants, the two methods produced no significant difference in hypertrophy when volume and effort matched. The authors also noted limits in the sensitivity of their hypertrophy measurements, so the null result does not prove biological equivalence. It does, however, offer no evidence that clusters are a muscle-building upgrade when the work is otherwise the same.

Advanced trainees are the evidence gap

One provocative ACSM finding was that training experience had minimal differential impact on strength and hypertrophy outcomes. Taken literally, it challenges the old tiered model that assigned increasingly complex progression methods to intermediate and advanced lifters.

But the stand also reports that its evidence base leaned heavily toward inexperienced trainees. That limitation changes how much confidence an experienced lifter should place in the apparent null result. Novices can grow under many competent programs because almost any unfamiliar, progressively demanding resistance work is a large new stimulus. Whether well-trained athletes gain the same results from simple progression as from carefully periodized long-term training remains unsettled. The absence of a demonstrated advantage is not proof that none exists in a population the literature has barely sampled.

Cardio does not automatically eat your gains

The "interference effect" is real under some combined-training setups, especially when endurance and resistance work pile up at high volumes with little recovery. It is not a universal property of doing cardio and lifting in the same week.

In a 2026 randomized trial, Conceição et al. assigned 19 previously untrained young men to 16 weeks of resistance training alone or concurrent resistance training plus long-interval HIIT. The concurrent protocol did not impair skeletal-muscle protein synthesis or hypertrophy relative to resistance training alone. This result applies to that HIIT protocol and population; it does not answer whether high-volume cycling, steady-state endurance work, or sport-specific training produces the same outcome.

A 2024 systematic review and meta-analysis by Huiberts et al. found that sex and current training status moderated concurrent-training adaptations. In middle-aged and older adults, a 2022 meta-analysis by Khalafi et al. found concurrent training preserved or improved cardiorespiratory fitness and muscular strength compared with either modality alone. Different populations, training volumes, interval types, and recovery windows help explain why "cardio kills gains" has never been a useful prescription.

BFR and metabolic claims need narrower language

BFR is another case where efficacy and superiority are different questions. The ACSM review found no consistent evidence that BFR outperforms conventional resistance training for hypertrophy or strength in healthy adults. Yet a randomized trial by Wang et al. found that low-load BFR produced fiber type-independent hypertrophy and improved functional capacity in older people. Those findings support BFR as an option when heavy loading is unsuitable; they do not establish BFR as better for healthy younger adults.

Resistance training also has health benefits beyond muscle size. In adults 65 and older, a 2024 meta-analysis of 28 studies found significant increases in muscle size (standardized mean difference 0.34, 95% confidence interval 0.16 to 0.52) and fiber area (0.54, 0.24 to 0.84) after resistance training. A review by de Santana et al. did not find a significant change in leg lean mass, a broader body-composition measure. In older women with sarcopenia, a 2026 meta-analysis of 12 randomized trials involving 518 participants found improved handgrip strength, gait speed, and knee-extension strength, with reported standardized mean differences of 0.43, 0.37, and 0.85, respectively.

More muscle does not automatically translate into proportionally better glucose control. A 2024 systematic review and meta-analysis by Paquin et al. found a complex, non-linear relationship between changes in fat-free mass—a surrogate for hypertrophy—and resistance-training improvements in glucose homeostasis. Muscle gain may contribute, but it is not a direct meter of metabolic benefit.

What to do with this evidence

• Choose loads you can train consistently: at least 30% 1RM can build muscle, while heavier loads remain more specific to maximal strength.
• Take lighter-load sets close to failure; do not assume heavy sets must reach failure.
• Prioritize progressive training, adequate weekly work, and repeatable effort before buying a more elaborate set structure.
• Keep cardio if it serves your health or performance goals, then manage total volume and recovery instead of fearing a categorical interference effect.

The 2026 ACSM review did not prove that every program works equally well. It showed that many claims of superiority have not survived the level of evidence needed to justify them—especially in novices, where most of the research has been done.

For experienced lifters, the sensible response is disciplined skepticism. Use simple progression as the default, retain complexity when it solves a real problem, and avoid treating a novice-heavy evidence base as the final word on advanced training.