Pre-discharge exercise tests were performed in 359 survivors of an acute myocardial infarction to determine the frequency, characteristics and prognostic implications of silent ischaemia. Tests were negative in 152 patients (42%), silent ischaemia was observed in 103 (29%) and painful ischaemia in 82 (23%). Heart rates at the development of ischaemia and the final double products were similar in both ischaemic groups but patients with silent ischaemia were able to exercise for longer (13.1 +/- 0.5 min) than those with painful ischaemia (9.3 +/- 0.5 min; P less than 0.0001). The 12 month mortality rose from 2% in patients with a negative test, to 4% in those with silent ischaemia and to 8% in those with painful ischaemia. Re-infarction rates increased similarly across the groups (3%, 8% and 18% respectively). Patients with silent ischaemia subsequently developed angina more frequently (47%) than those with negative tests (16%; P less than 0.001). These results suggest that exercise-induced silent ischaemia following myocardial infarction was common, occurring in 29% of patients. Although the final myocardial oxygen consumption was similar in both ischaemic groups those with silent ischaemia were able to exercise for longer. Finally silent ischaemia conferred an intermediate risk of death or re-infarction and was a strong predictor of subsequent angina pectoris.