Clinical, exercise, and angiographic variables, and long-term follow-up were compared in patients, who, during maximal Bruce exercise testing after a first acute myocardial infarction (AMI), had positive responses to exercise testing (n = 116, 38% of 303) with (n = 23, group I) or without (n = 93, group II) angina. Group I patients more often (52 vs 19%, P < 0.001) had a history of pre-infarction angina. Group II had a greater proportion (75 vs 52%, P < 0.05) of inferior wall AMI, whereas group I had a greater proportion (30 vs 19%, P < 0.01) of non-Q wave AMI. Total exercise duration was significantly (P < 0.01) longer in group II (7.6 +/- 3.2 vs 5.5 +/- 3.1 min). Maximal exercise heart rate (144 +/- 22 vs 133 +/- 21, beats.min-1 P < 0.05) was also higher in group II. A greater proportion of group II patients (37 vs 9%, P < 0.05) had single-vessel disease, whereas multivessel disease was more common (91 vs 63%, P < 0.03) in group I. Left ventricular function was similar in both groups. During follow-up (48 +/- 22 months) the incidence of cardiac death (group I, 3.3%, group II, 4.8%), of recurrent infarction (group I, 4.8%, group II 3.3%), and of revascularization procedures (group I, 28.5%, group II, 19.8%) were similar in both groups. Although asymptomatic exercise-induced ischaemia was associated with better exercise performance and less extensive coronary disease than symptomatic ischaemia, it had the same long-term prognostic implications.