The role of exercise radionuclide angiography in the therapeutic strategy of chronic aortic insufficiency remains controversial. In order to assess the value of this technique, 47 patients with pure chronic aortic insufficiency were evaluated before and one year after valve replacement. The preoperative EF decreases or does not increase on exercise (51.9% vs 48 +/- 10%; p < 0.001) in the majority of patients (78%), whereas, after surgery, the EF tends to increase on exercise (55 +/- 11% vs 57 +/- 11%; NS). Despite optimal surgical correction, 16 of the surviving 46 patients still had left ventricular dysfunction at one year after surgery. The preoperative parameters correlating with this dysfunction were, in order, resting radionuclide EF (r = 0.65; p = 0.0001) and the echocardiographic parameters: left ventricular endosystolic dimension, fractional shortening, and the radius/thickness ratio. The preoperative exercise radionuclide parameters did not correlate with postoperative left ventricular dysfunction. In patients without postoperative left ventricular dysfunction, the EF increased on exercise, contrary to the other group. The exercise EF remains valuable for retrospective evaluation of surgical benefit. The exercise radionuclide EF does not reflect myocardial contractile reserve alone but also the conditions of left ventricular load. This study confirms the superiority of the resting radionuclide EF and echocardiographic parameters over exercise radionuclide EF for the prediction of postoperative left ventricular dysfunction.