To assess inter- and intraobserver variation in the echocardiographic diagnosis of mitral valve prolapse, three independent observers analyzed M-mode echocardiograms (n = 80) and two-dimensional echocardiograms (n = 65) of patients with a mobile midsystolic click with or without a late or holosystolic murmur. In addition, a control group of 100 normal echocardiograms were interspersed among the echocardiograms of patients with mitral valve prolapse and were then interpreted. Each of the three observers analyzed all M-mode and two-dimensional echocardiograms initially and then 2 weeks later for the presence or absence of mitral valve prolapse. M-mode echocardiographic criteria for mitral valve prolapse consisted of late systolic posterior motion (greater than or equal to 3 mm) of one or both mitral leaflets or holosystolic hammocking (greater than or equal to 3 mm) of one or both mitral leaflets. Two-dimensional echocardiographic criteria for mitral valve prolapse consisted of: posterior systolic arching of one or both mitral leaflets in the parasternal long-axis view, and/or posterior systolic bowing of one or both mitral leaflets in the apical four-chamber view posterior to the plane of the mitral anulus, and/or excessive posterior coaptation of the mitral leaflets in either view flush with or posterior to the plane of the mitral anulus. There was insignificant observer variation both in the M-mode and two-dimensional echo groups, as determined using Cochran's Q test.(ABSTRACT TRUNCATED AT 250 WORDS)