To test the hypothesis that the combined use of the time-varying elastance concept and conventional circumferential stress-shortening relations would elucidate differential mechanisms for left ventricular systolic dysfunction in severe, chronic aortic regurgitation and therefore predict the functional responses to aortic valve replacement, 31 control patients and 37 patients with aortic regurgitation were studied. The studies included micromanometer left ventricular pressure determinations, biplane contrast cineangiograms under control conditions and radionuclide angiograms under control conditions and during methoxamine or nitroprusside infusions with right atrial pacing. The patients with aortic regurgitation were classified into three groups: Group I had normal Emax and stress-shortening relations, Group II had abnormal Emax but normal stress-shortening relations and Group III had abnormal Emax and stress-shortening relations. The left ventricular end-diastolic and end-systolic volumes showed a progressive increase and the ejection fraction showed a progressive decrease from Group I to III; these values differed from those in the control patients (p less than 0.001). In Group I, there was a decrease in left ventricular volumes (p less than 0.05) but no significant change in ejection fraction (61 +/- 7% versus 63 +/- 4%) after aortic valve replacement. In contrast, in Group II, reduction in left ventricular volumes (p less than 0.01) was associated with an increase in ejection fraction from 50 +/- 8% to 64 +/- 11% (p less than 0.01). Finally, in Group III, reduction in left ventricular volumes (p less than 0.05) was associated with a further decrement in ejection fraction from 35 +/- 13% to 30 +/- 13%. Group I patients had compensated adequately for chronic volume overload. However, Group II had left ventricular dysfunction that was associated with an increase in the left ventricular volume/mass ratio compared with that in the control patients and Group I (p less than 0.05 for both), suggesting inadequate hypertrophy and assumption of spherical geometry. Finally, irreversible myocardial dysfunction had supervened in Group III. In conclusion, a combined analysis of left ventricular chamber performance using the time-varying elastance concept and myocardial performance using conventional circumferential stress-shortening relations provides complementary information that elucidates differential mechanisms for left ventricular systolic dysfunction and therefore predicts the functional response to aortic valve replacement.