Serial echocardiographic analyses of left ventricular hypertrophy and function, with validation of extent of shortening by first pass radionuclide angiography, was performed in 16 patients before and after surgical correction of severe aortic valve regurgitation. All patients were symptomatic (predominantly in New York Heart Association functional class III or IV) before operation but were in class I or II after operation. The preoperative pattern of eccentric hypertrophy (increased mass with normal ratio of left ventricular cross-sectional wall area to cavity area) changed immediately after operation to a pattern of concentric hypertrophy (increased mass with increased ratio of left ventricular cross-sectional wall area to cavity area) because of a significant reduction in chamber size and increase in wall thickness. On late follow-up (9 to 35 months, average 15 months after operation), the hypertrophy lessened significantly, the cross-sectional area of the ventricular wall decreasing to 21.1 ± 5.4 (mean ± standard deviation) cm2 from a preoperative average of 31.6 ± 4.8 cm2 (P < 0.01), and the ratio of wall area to cavity area was once again normal. In the same period, left ventricular enddiastolic diameter decreased from 6.52 ± 0.68 to 4.64 ± 0.52 cm (P < 0.01). Preoperatively, ejection phase indexes were normal or only marginally depressed in 12 of 16 patients but were moderately depressed in the remaining 4. At early follow-up (average 4 months) ventricular shortening tended to increase; and at late follow-up the fractional shortening of the minor axis, the ejection fraction and the mean velocity of circumferential fiber shortening increased to 0.39 ± 0.07, 0.68 ± 0.10 and 1.26 ± 0.22 circumference/sec, respectively, from preoperative values of 0.33 ± 0.09, 0.60 ± 0.14 and 1.05 ± 0.31 circumferences/sec (P < 0.05 for each index). In the four subjects with preoperative depression of left ventricular function, the extent and speed of myocardial shortening at late follow-up became normal in three subjects and remained moderately depressed in one patient. Paradoxical septal motion was observed immediately postoperatively and in the early follow-up studies, but it was noted in only 3 of 16 cases by the late follow-up period. Provided septal dyskinesia was not present, echocardiographic and first pass radionuclide determinations of ejection fraction correlated highly (r = 0.92). It is concluded that when aortic valve replacement for symptomatic aortic regurgitation is undertaken prior to severe myocardial decompensation, improvement in clinical status is associated with significant regression of myocardial hypertrophy, reduction in left ventricular size, evolution of a normal mass volume ratio, recovery of septal dyskinesia as revealed on echocardiography, and improvement in left ventricular function. These data do not define the type and degree of left ventricular dysfunction which is irreversible.