Elevated plasma natriuretic peptide levels after AMI (acute myocardial infarction) are associated with adverse outcome. The aim of the present study was to examine the relationship of plasma N-ANP (N-terminal pro-atrial natriuretic peptide) and N-BNP (N-terminal pro-B-type natriuretic peptide) with mortality and heart failure following AMI. We studied 403 patients with AMI. Blood was sampled for measurement of N-ANP and N-BNP on a single occasion between 72 and 96 h after symptom onset. Natriuretic peptide levels were related to all-cause mortality and heart failure episodes. During follow up (median, 462 days; range 5-764), 43 (10.7%), 25 (6.2%) and 49 (12.2%) patients suffered death, heart failure hospitalization and outpatient heart failure respectively. Only N-BNP (P < 0.0005), N-ANP (P = 0.005) and previous AMI (P = 0.016) were independently predictive of death. N-BNP, but not N-ANP, predicted 30-day mortality. N-ANP, but not N-BNP, predicted mortality after 30 days. N-BNP was the better predictor of heart failure. N-ANP and N-BNP were above the median in 35 and 38 respectively, of 43 patients who later died. N-ANP, N-BNP, or both were above the median in 41 out of 43 patients. Of 25 patients hospitalized with heart failure, N-ANP and N-BNP was above the median in 20 and 24 respectively, and one or other was elevated in all cases. Above-median N-ANP predicted 36 and N-BNP predicted 41 out of 49 episodes of outpatient heart failure. One or other peptide was above the median in 45 out of 49 patients. Our results indicate that N-BNP predicts 30-day and N-ANP >30-day mortality. We conclude that consideration of both N-ANP and N-BNP identifies a greater number of patients at risk of death or heart failure than either peptide alone.