Twenty-four patients with unstable angina pectoris, defined as chest pain at rest with transient ST segment deviation of at least 1 mm, were randomly assigned to blinded treatment with either placebo or intravenous recombinant human tissue-type plasminogen activator (rt-PA). Before randomization, all patients were treated with oral 13-blockers, calcium antagonists, nitrates, and continuous intravenous heparin for a monitoring period of 12 to 28 hr. After this monitoring period the 24 patients were randomly assigned to receive either placebo or 1.75 mg/kg intravenous rt-PA given over 12 hr at a rate of 0.75 mg/kg over 1 hr, 0.5 mg/kg over 4 hr, and 0.5 mg/kg over 7 hr. One patient, assigned to receive placebo, developed acute myocardial infarction after randomization but before receiving the study drug. Ischemic events were recorded during a hospital follow-up period of at least 4 days unless a further intervention was indicated or the coronary angiogram was normal. The follow-up period was 7 5 days (mean + SD) after the placebo infusion and 8 ± 4 days after the infusion of rt-PA. Unstable angina pectoris persisted after the completion of the infusion in six of 11 patients receiving placebo and only one of 12 patients receiving rt-PA (p < .03). Coronary angiography, performed 38 ± 19 hr after the infusion, demonstrated subocclusive thrombus in eight of 11 patients receiving placebo but in none of 11 patients treated with rt-PA (p < .002). One patient on rt-PA refused coronary angiography. Persistence of unstable angina pectoris correlated significantly with the presence of intracoronary thrombus (p < .003). The infusion of rt-PA caused a decrease of fibrinogen to 66 ± 19% after 1 hr, to 65 ± 21% after 5 hr, and to 64 ± 29% after 12 hr. Oozing at puncture sites was not observed in patients receiving placebo but was seen in eight of 12 patients treated with rt-PA. In three patients on rt-PA, bleeding complications necessitated termination of the infusion within 5, 8, and 11 hr after onset. Thus, in this group of patients with unstable angina pectoris, intravenous infusion of rt-PA appeared to lyse subocclusive intracoronary thrombi. The removal of intracoronary thrombus by rt-PA was associated with reduction of the severity of resting angina. The decision for further intervention could then be made on the basis of residual coronary occlusion.