The prognostic value of preoperative echocardiographic data was assessed in 32 patients who underwent aortic valve replacement for chronic aortic regurgitation. All patients had preoperative studies and were followed up prospectively for 1 to 6 years after surgery. Postoperatively, 25 patients (Group A) achieved a normal left ventricular end-diastolic dimension and a significant regression of myocardial hypertrophy; 7 patients (Group B) had persistent left ventricular enlargement. During the follow-up period, the patients in Group A had fewer symptoms and used fewer medications than those in Group B. Moreover, survival at 4 years appeared to be better in Group A (96%) than in Group B (71%); two patients in Group B died with congestive heart failure; there were no such deaths in Group A. Preoperatively, a left ventricular dimension at end-diastole (DED) larger than 3.8 cm/m2 body surface area, a dimension at end-systole (DES) greater than 2.6 cm/m2 body surface area, an end-diastolic radius/wall thickness ratio (R/Th) greater than 3.8 or a product of R/Th and left ventricular systolic pressure (P X R/Th) exceeding 600 are predictive of a Group B result. If end-systolic dimension is greater than 2.6 and P X R/Th is greater than 600, all Group B patients can be identified; all but one patient in Group A had an end-systolic dimension less than 2.6 and P X R/Th less than 600. It is concluded that patients with chronic aortic regurgitation who are at risk of persistent postoperative left ventricular enlargement (with associated cardiac symptoms and reduced survival) can be identified by preoperative echocardiography.