We evaluated the value of color and pulsed Doppler transesophageal echocardiographic parameters and of V waves in estimating the severity of mitral regurgitation (MR) in 62 consecutive patients (38 men and 24 women, aged 39 to 80 years) with angiographically proven chronic pure MR (15 grade I/II, 47 grade III/IV). Twenty patients were examined before cardiac surgery under general anesthesia. Sensitivity, specificity, and positive and negative predictive values of systolic pulmonary venous flow reversal for the presence of grade III/IV MR were 87%, 93%, 98%, and 64%, respectively, these were for jet areas > or = 8.0 cm2--66%, 100%, 100%, and 48%, for jet lengths > or = 50 mm--70%, 87%, 94%, and 48%, for enlarged V waves--86%, 38%, 83%, and 43%, and for either flow reversal or a jet area > or = 8.0 cm2--96%, 93%, 98%, and 88%. We conclude that a combination of measurements improved the negative predictive value considerably, which is of importance in a population with a high pretest probability of severe MR. Enlarged V waves are not reliable in predicting severe MR. The optimal cutoff value for jet area and jet length was lower in anesthesized patients than in conscious patients; in anesthesized patients, sensitivity, specificity, and positive and negative predictive values of jet area > or = 5.0 CM2 for grade III/IV MR were 67%, 100%, 100%, and 50%, respectively; these were 87%, 100%, 100%, and 71% for flow reversal. Because the results of mitral repair are often evaluated with transesophageal echocardiography during surgery, our findings have clinical implications for evaluation of severe MR in anesthesized patients: pulmonary venous flow direction is the first-choice measure; jet area can be used when a low cutoff point is chosen.