The independent predictive role of ventricular premature complex (VPC) frequency in the stratification of mortality risk after acute myocardial infarction (AMI) was established in the prethrombolytic era by extensive multicenter trials. Thrombolysis has lead to important changes in the natural history of patients after AMI, so that reassessment of established risk factors is now required. The prognostic significance of VPCs was assessed in 680 patients, of whom 379 received early thrombolytic therapy. All patients underwent 24-hour Holter monitoring in a drug-free state between 6 and 10 days after AMI. Patients were followed up for 1 to 8 years. During the first year of follow-up, cardiac death occurred in 33 patients, sudden death in 24, and sustained ventricular tachycardia in 20. Mean VPC frequency was significantly higher in patients who died of cardiac causes, in those who died suddenly, and in those with arrhythmic events during the first year of follow-up. This was also true when patients who did and did not undergo thrombolysis were considered separately. The positive predictive accuracy of VPC frequency in predicting adverse cardiac events was greater in patients who did than did not undergo thrombolysis. At a sensitivity level of 40%, the positive predictive accuracy for cardiac mortality and arrhythmic events for the group with thrombolysis was 19.4% and 25.8%, respectively, compared with 16% and 16% for those without thrombolysis. Moreover, the highest VPC frequency for the dichotomy of patients into high-and low-risk groups was 25 VPCs/hour for patients without thrombolysis. VPC frequency appears to be more highly predictive of prognosis after AMI in patients who have undergone thrombolysis than in those who have not, but the optimal frequency for dichotomy is higher in the former.