Thrombolysis has been reported to restore coronary blood flow in patients with acute myocardial infarction (AMI). However, the relationship between fibrinolytic treatment and evidence of myocardial reperfusion has not been adequately assessed. Accordingly, we measured great cardiac vein blood flow (GCVF:thermodilution) in 12 patients (Group 1) presenting with AMI (chest pain < 4 hours and ST elevation in the anterior leads) before and following i.v. urokinase (UK:2 million U/90 min). Ten patients receiving conventional treatment served as controls (Group 2). UK induced a significant increase of GCVF (from 101 +/- 24 to 164 +/- 42 ml/min, p < 0.001). Maximal increase occurred after 50 +/- 54 min from drug infusion. Conversely, changes in GCVF were not significant in Group 2 (from 103 +/- 35 to 106 +/- 31 ml/min, NS). Following 24 hours changes in GCVF were still consistent only in Group 1 patients. Individual analysis during 24 hours showed marked fluctuations of GCVF peak values in Group 1 patients (62 +/- 43%), but not in Group 2 (29 +/- 21%). Thus, UK induces a marked increase of GCVF in most patients with anterior AMI; such increase suggests that reperfusion occurs early (i.e. within 1 hour) from UK administration. Fluctuations of GCVF during monitoring are magnified by thrombolysis, suggesting intermittent coronary reocclusion in the early hours of AMI.