Rescue percutaneous transluminal coronary angioplasty (PTCA) has been used to establish reperfusion after failed thrombolysis, and the goal of this study was to examine the angiographic and clinical outcomes after rescue PTCA performed for an occluded artery 90 minutes after thrombolysis. Four hundred two patients with acute myocardial infarction were randomized to receive either anistreplase (APSAC), recombinant tissue plasminogen activator, or their combination in the Thrombolysis in Myocardial Infarction (TIMI) 4 trial. The angiographic and clinical outcomes of patients with a patent artery 90 minutes after thrombolysis were compared with those of patients with an occluded artery treated in a nonrandomized fashion with either rescue or no rescue PTCA. At 90 minutes, the number of frames required to opacify standard landmarks (corrected TIMI frame count) was significantly lower (i.e., flow was faster) after successful rescue PTCA (27 +/- 11) than that in patent arteries after successful thrombolysis (39 +/- 20, p < 0.001), and the incidence of TIMI grade 3 flow was correspondingly higher after successful rescue PTCA (87% vs 65%, p = 0.002). In-hospital adverse outcomes (death, recurrent acute myocardial infarction, severe congestive heart failure, cardiogenic shock or an ejection fraction <40%) occurred in 29% of successful rescue PTCAs and in 83% of failed rescue PTCAs (p = 0.01). Among all patients in whom rescue PTCA was performed (successes and failures combined), 35% of patients experienced an adverse outcome, which was the same as the 35% incidence observed in patients not undergoing rescue PTCA (p = NS) and tended to be higher than the 23% incidence observed in patients with patent arteries (p = 0.07). Although successful rescue PTCA for an occluded artery at 90 minutes results in restoration of flow that is superior to that of successful thrombolysis, the incidence of adverse events for the strategy of rescue PTCA as a whole was the same as that of undertaking no PTCA.