The absence of controlled evidence and the high prevalence of mitral-valve prolapse have created substantial uncertainty about whether this condition is an important risk factor for bacterial endocarditis. We evaluated this risk in a case-control study of hospital inpatients who had undergone echocardiography and who lacked any known cardiovascular risk factors for endocarditis, apart from mitral-valve prolapse and isolated mitral-regurgitant murmurs. Thirteen (25 per cent) of 51 patients with endocarditis had mitral-valve prolapse, as compared with 10 (seven per cent) of the 153 matched controls without endocarditis. For the 51 matched case-control sets, the odds ratio (8.2; 95 per cent confidence interval, 2.4 to 28.4) indicated a substantially higher risk of endocarditis for people with mitral-valve prolapse than for those without it. This association remained statistically significant when parenteral drug abuse and routine antibiotic prophylaxis preceding dental work and other forms of instrumentation were taken into account. Furthermore, the risk may be higher than is indicated by this study, since 46 per cent of the controls underwent echocardiography for clinically suspected mitral-valve prolapse, suggesting an overrepresentation of mitral prolapse in the control group. The results support the contention that mitral-valve prolapse is a significant risk factor for bacterial endocarditis.