ObjectiveTo evaluate the clinical and basic science evidence surrounding the hypothesis that stretching immediately before exercise prevents injury.Data Sources and SelectionMEDLINE was searched using MEDLINE subject headings (MeSH) and textwords for English- and French-language articles related to stretching and muscle injury. Additional references were reviewed from the bibliographies, and from citation searches on key articles. All articles related to stretching and injury or pathophysiology of muscle injury were reviewed. Clinical articles without a control group were excluded.ResultsThree (all prospective) of the four clinical articles that suggested stretching was beneficial included a cointervention of warm-up. The fourth study (cross-sectional) found stretching was associated with less groin/buttock problems in cyclists, but only in women. There were five studies suggesting no difference in injury rates between stretchers and nonstretchers (3 prospective, 2 cross-sectional) and three suggesting stretching was detrimental (all cross-sectional). The review of the basic science literature suggested five reasons why stretching before exercise would not prevent injuries. First, in animals, immobilization or heating-induced increases in muscle compliance cause tissues to rupture more easily. Second, stretching before exercise should have no effect for activities in which excessive muscle length is not an issue (e.g., jogging). Third, stretching won't affect muscle compliance during eccentric activity, when most strains are believed to occur. Fourth, stretching can produce damage at the cytoskeleton level. Fifth, stretching appears to mask muscle pain in humans.ConclusionThe basic science literature supports the epidemiologic evidence that stretching before exercise does not reduce the risk of injury.