To compare the non-invasive methods of quantification of acute myocardial infarction (AMI) [two dimensional echocardiography (2DE), standard 12-leads ECG, and enzymatic indices as MB-CK peak activity and MB-CK time activity curve expressed by an extension index (EI-MBCK)] in relation to their prognostic value, 79 patients with a first AMI were evaluated. We have observed in a three months follow-up a total mortality of 12.6%. The infarct size, calculated echocardiographically by a segment score, was correlated with the number of pathological Q waves in the standard ECG (rho= 0.83). Peak MB-CK enzyme and EI-MBCK correlated both with the segment score, but with a lower correlation coefficient (rho= 0.67). To identify patients at different risk, discriminant analysis was used which gave the following limit values for the patients at a very high risk: 2DE score = 17; number of Q waves = 7; peak MB-CK = 176 U/L; EI-MBCK = 54 grEq/m2; for the patients at a very low risk: score = 6; number of Q waves = 2; peak MB-CK = 35; EI-MBCK = 15. To verify if the association of these different techniques could improve the predictivity, a discriminant bivariate function analysis with three variables was calculated. The resulting equation was: Z = 2.31 X 2DE score + 8.59 X number of Q waves - 0.23 X peak MB-CK. Changing peak MB-CK value with EI-MBCK did not improve the statistical significativity. The results have confirmed that the integration of all the three variables improved the prognostic predictivity. According to the risk Z obtained, the patients were allocated into classes of different risk: values of Z greater than 57 or less than 18 could identify patients respectively at a very high or at a very low risk. For values between 37 and 42 the prognosis remains uncertain. Among the three variables, 2DE and ECG showed an equivalent prognostic accuracy, whereas enzyme indices had a lower prognostic influence, especially in the presence of large infarcts. Thus, 2DE, ECG and enzyme indices can identify patients at increased risk; the individual method seems to be inadequate; to obtain valid predictive informations it is necessary to integrate all the three non invasive techniques.