Delayed-onset muscle soreness

What it is

How reviews and consensus statements define it

The 2013 Munich consensus statement on muscle injuries in sport calls delayed-onset muscle soreness, soreness that starts hours to days after exercise, type 1B¹. That places it among functional muscle disorders from overexertion¹. The statement describes it as more generalised muscle pain after unaccustomed eccentric movements that slow the body down¹.

The same statement sorts functional disorders, which show no visible tear, apart from structural injuries, which do; the muscle strain page covers tears¹.

What causes it is not settled

A 2018 narrative review, a summary written without a set, checkable search method, says the exact pathway of the soreness is still unknown⁶. It names damage to muscle cells from unfamiliar or eccentric exercise as the main mechanism under current thinking⁶. A 2003 narrative review lists up to six proposed mechanisms, including muscle damage, inflammation and lactic acid⁷. It concludes that two or more of them together likely explain the soreness⁷.

How it comes on and how long it lasts

Sources differ a little on timing^2,6:

Source	What it reports
A 2019 review	Appears 12 to 24 hours after unaccustomed eccentric exercise, peaks at 24 to 72 hours, gone within 5 to 7 days²
A 2017 Cochrane review	Peaks 24 to 72 hours after the first bout³
The 2018 narrative review	Peaks 48 to 72 hours after exercise⁶
The 2013 Munich statement	Settles on its own, usually within a week¹

A Cochrane review is a systematic review done to the Cochrane Collaboration’s published standards. The Munich statement describes inflammatory pain, pain at rest, and muscles that feel stiff and weak¹. The 2018 narrative review lists lower force, painful limits on movement, stiffness, swelling and poor function of nearby joints⁶. The 2019 review in the table also notes that how sore a muscle feels matches other markers of muscle damage poorly².

Who it affects

Athletes at every level

The 2003 narrative review calls it familiar to both elite and novice athletes⁷. The 2018 narrative review calls it a mild injury yet one of the most common reasons for reduced sports performance⁶.

When it tends to happen

Reviews tie it to exercise that is intense, unaccustomed or eccentric^3,5. The 2003 narrative review states that how hard and how long people exercise both matter to its onset⁷. The same review says it is most common at the start of a sporting season, after a spell of less activity⁷.

Men, women and the menstrual cycle

A 2020 systematic review of 23 trials found no significant difference in soreness between men and women after one bout of maximal eccentric exercise⁸.

A 2021 meta-analysis of 19 articles studied women with regular menstrual cycles⁹. Soreness and strength loss differed between phases of the cycle⁹. Creatine kinase, a blood marker of muscle damage, did not differ between phases⁹.

Soreness reported in training studies

Some exercise studies recorded soreness as an outcome^10,11,12:

Group studied	What was found
40 men doing the Nordic hamstring curl at four doses for six weeks	The group that went from low to high volume had more soreness than the others¹⁰
57 male football players in a randomised trial	More soreness with the Copenhagen adductor plank than with an adductor squeeze¹¹
10 athletes doing free sprints, or sprints towing a sled of 10% or 20% of body mass	All three raised soreness of the muscles that straighten and bend the knee for up to 72 hours¹²

What studies found

One bout protects against the next

The repeated bout effect is the smaller damage and soreness seen when the same unaccustomed exercise is done a second time. A review on the eccentric training page describes less soreness and swelling after the repeat². It says the protection appears after one bout, but seems more complete after several sessions².

A 2023 meta-analysis of 23 articles studied people with no recent resistance training⁴. One earlier eccentric or isometric bout was followed by less soreness 24 to 96 hours after a later damaging bout than no earlier bout⁴. Comparing studies without pooling them, the same review found an earlier bout 2 to 4 days ahead worked best⁴. It worked less well right before, or 1 to 3 weeks before⁴. Eccentric, higher-volume, harder and more lengthened contractions gave more protection⁴.

In a 2001 study, 35 male students repeated eccentric exercise of the elbow flexors, which bend the elbow, after 6, 9 or 12 months¹³. Soreness was lower after the second bout in the 6-month group¹³. The authors concluded the effect lasts at least 6 months for most measures, but is lost between 9 and 12¹³.

What a 2025 umbrella review found

The 2025 umbrella review covered 29 systematic reviews of 863 randomised trials and 24 approaches used after exercise in healthy adults⁵. Seventeen of the reviews were of critically low quality, and only two were rated high⁵. It found some strong evidence, at various time points, for six approaches⁵. These were cooling, cryostimulation (cold exposure), contrast therapy (alternating hot and cold), massage, light therapy and kinesiology tape⁵. Evidence for stretching, exercise and electrical stimulation was weak⁵.

A 2018 meta-analysis of 99 studies found small to large drops in soreness with six methods¹⁴. These were massage, active recovery, cryotherapy (cold therapy), compression garments, contrast water therapy and immersion¹⁴. The 2003 narrative review had reported no effect on soreness from cold, stretching, ultrasound or electrical current⁷. Both newer reviews differ from it on cold, reporting less soreness with cold-based methods^5,14.

Cold water and cold air

A 2012 Cochrane review pooled 17 small trials in 366 people, and rated study quality low¹⁶. It found less soreness with cold-water immersion than with rest or no intervention at 24, 72 and 96 hours, though results varied widely¹⁶.

Two newer reviews also found less soreness with cold water at 24 hours, but not in every case^17,18:

A 2022 meta-analysis of 52 studies in active people found cold water lowered soreness 24 hours after high-intensity exercise, against passive recovery¹⁷. After eccentric exercise, it improved only muscle power at 24 hours¹⁷.
A 2026 network meta-analysis of 51 trials (1,243 people) found no cold method lowered soreness right after exercise¹⁸. Cold water lowered it at 1 and 24 hours, and local cold had the largest effect at 48 hours¹⁸.
A 2015 Cochrane review pooled four lab trials of whole-body cryotherapy, exposure of the whole body to extreme cold, in 64 active adults¹⁹. Its very low-quality evidence was too little to tell whether it lowered self-reported soreness compared with rest¹⁹.

Massage, compression and rolling

A 2020 meta-analysis of 29 randomised studies (1,012 people) linked massage with small but significant improvements in soreness and flexibility¹⁵. It found no evidence of gains in strength, jumping, sprinting, endurance or fatigue¹⁵. Massage lowered pain or soreness by 13%, but one outlying study made results vary so much that the true size of any benefit is uncertain¹⁵. The 2018 meta-analysis of 99 studies found massage the most effective method it examined for soreness and felt fatigue¹⁴.

On compression garments, tight-fitting clothing, a 2014 meta-analysis of 12 studies found a moderate effect on how severe soreness was²⁰. A 2026 meta-analysis of nine trials in 158 adult endurance athletes found soreness at about 24 hours favoured compression, at very low certainty²¹.

On foam rolling, one 2024 meta-analysis found lower soreness and one 2023 meta-analysis found no significant difference^22,23.

Stretching

A 2011 Cochrane review of 12 studies found stretching before or after exercise gave no clinically important drop in soreness in healthy adults²⁴. A 2021 meta-analysis of 10 trials in 229 people agrees, finding no effect of stretching after exercise on soreness at 24 to 72 hours²⁵. Confidence in that evidence was very low²⁵.

Anti-inflammatory drugs

A 2018 meta-analysis of 41 human and animal studies looked at NSAIDs, drugs such as ibuprofen that ease pain and inflammation²⁶. After acute muscle injury, it found a small to medium drop in strength loss, soreness and creatine kinase taken together²⁶. A 2024 meta-analysis of NSAIDs around resistance exercise disagrees on strength and creatine kinase²⁷. Strength was similar to placebo at 0 and 24 hours, lower with NSAIDs at 48 hours, and creatine kinase was unchanged²⁷. Its authors judged NSAIDs ineffective for performance, strength and recovery²⁷.

A 2024 randomised crossover trial, in which each person tried every option in turn, studied 12 healthy adults aged 18 to 42²⁸. One dose of each of three NSAIDs, taken before plyometric (jumping) exercise, made no significant difference to soreness against a placebo²⁸. Its Crossref record, publisher-supplied data, lists the Medical Research and Materiel Command and the Military Operational Medicine Research Program as funders²⁸. The authors declare no conflicts of interest²⁸.

Supplements

The three newer vitamin reviews agree with the 2017 Cochrane review in finding no clear soreness benefit from antioxidant vitamins^3,29,30,31.

Supplement	Review	What it found
Antioxidants	2017 Cochrane review, 50 placebo-controlled trials, 1,089 people (88% male)	A small difference in soreness favouring high doses, well below the smallest change people notice; moderate to low-quality evidence³
Vitamin C	2020 meta-analyses, 18 placebo-controlled trials, 313 healthy volunteers	No effect on soreness, strength or creatine kinase after one bout of exercise²⁹
Vitamin E	2024 meta-analysis, 20 placebo-controlled trials, 298 healthy people	No effect on soreness, strength or creatine kinase; results for soreness varied widely³⁰
Vitamins C and E	2021 systematic review, 14 randomised trials, 280 healthy people aged 16 to 30	Soreness lower than placebo at every time point in only 3 studies; too little evidence to confirm a benefit³¹
Flavonoid polyphenols (a group of plant compounds)	2021 meta-analysis, 26 placebo-controlled trials	Taken for 6 or more days, 4.12% less soreness and 7.14% better strength recovery over 96 hours; creatine kinase unchanged³²
Branched-chain amino acids	2024 meta-analysis, 18 randomised trials in healthy active people	Lower soreness at 24 to 96 hours, but not right after exercise³³
Curcumin	2022 meta-analysis, 10 trials	Lower soreness ratings; the authors call for larger, well-designed studies³⁴
Omega-3	2024 systematic review, 13 controlled trials in healthy adults	Mixed results for soreness³⁵

Soreness and the return to sport

The 2003 narrative review states that the soreness may change the order and pattern in which muscles are recruited⁷. It suggests these changes may raise the risk of further injury if athletes return to sport too early⁷.

A 2024 systematic review of 20 studies in healthy adults and adolescents measured performance after muscle-damaging resistance and plyometric exercise³⁶. Sprint and change-of-direction performance were significantly worse for up to 72 hours, alongside raised creatine kinase and soreness³⁶. It reported performance, not injury, so it does not test the older review’s suggestion³⁶.

Connections

The muscle page covers how eccentric exercise damages muscle unused to it more than other exercise does². The eccentric training page covers the repeated bout effect in more depth³⁷. Tears are covered on the muscle strain page¹.

Soreness has been recorded in studies of the Nordic hamstring curl, the Copenhagen adductor plank and free or towed-sled sprinting^10,11,12.