ALTHOUGH the concept of coronary thrombolysis for the treatment of acute evolving myocardial infarction has been extant for over three decades,1 it is only in the last few years that profound benefits have been consistently documented in large numbers of patients, reflected by improved left ventricular function,2-5 decreased incidence and severity of congestive heart failure,5 decreased incidence of late fatal arrhythmias,6 and improved short-7-10 and long-term11,12 mortality. However, as with any medical or surgical treatment, pharmacologic activation of the fibrinolytic system entails a risk—in its case, the primary risk being that of bleeding. This risk is not a "side effect" but a direct consequence of the desired action when lytic agents are administered, ie, fibrin degradation and clot dissolution. Unfortunately, plasmin activation cannot be confined presently to pathological deposits of fibrin, and protective thrombi also will be disrupted. Nevertheless, spontaneous bleeding (gastrointestinal, genitourinary, retroperitoneal,