In order to evaluate the severity of aortic valve regurgitation (AVR) by means of simple criteria, we compared the feasibility and reliability of two methods: (1) pulsed doppler ultrasound suprasternal recording in the aortic sinus area, with calculation of the regurgitation fraction by planimetry of the systolic and diastolic curves, and with measurement of end-diastolic velocity, or end-diastolic doppler effect (EDDE); this was done in 114 subjects (84 patients with AVR and 30 controls); (2) continuous wave doppler ultrasound apical recording of the left intraventricular jet, with measurement of the velocity decrease slope (S) and of the velocity half-decrease time (T 1/2); this was done in 46 patients with AVR. Doppler results were compared with Seller's angiographic classification of AVR in 4 grades. Planimetry could be performed in only 41% of patients in this series. This measurement seems to be feasible only when perfect recording of an increased systolic flow (peak velocity higher than 1.2 m/s) can be performed, which is usually limited to cases with major regurgitation. EDDE was easier to record (84/84 patients). When above 5 cm/s it is a good reflection of AVR severity, and when above 20 cm/s it indicates a major AVR (3/4 or 4/4 at angiography), with an 81% sensitivity and a 91% specificity. Continuous wave doppler ultrasound apical recording could be used in 80% of the cases (37/46 patients). With this method, a more than 3 m/s slope is a highly specific (8/8) but not very sensitive (8/13) sign of major AVR. A T 1/2 value lower than 650 ms is a specific (12/12) and sensitive (12/13) sign of severe AVR.(ABSTRACT TRUNCATED AT 250 WORDS)