Early complications may hamper efforts to hasten discharge after primary percutaneous coronary intervention (PCI) for myocardial infarction (MI). Glycoprotein IIb/IIIa inhibitors, by reducing early recurrent ischemia, may aid in these efforts. We examined whether adjunctive abciximab could accelerate discharge and reduce costs within a trial of primary PCI after acute MI. The CADILLAC trial randomized 2,082 patients with MI to 1 of 4 reperfusion strategies in a 2 x 2 factorial design: angioplasty, angioplasty with abciximab, stent implantation, or stenting with abciximab. Patients randomized to abciximab had postprocedural heparin withheld, and discharge scheduled for days 1.5 to 2 (low-risk patients) or days 2 to 3 (high-risk patients) after MI if they were stable. Other patients were discharged at the physician's discretion. Abciximab treatment was associated with significant reductions in the primary end points of in-hospital death, reinfarction, ischemic target vessel revascularization (TVR), or disabling stroke (5.6% vs 2.7%, p = 0.003)--largely reflecting reduced ischemic TVR (3.8% vs 1.4%, p = 0.002)--and in early subacute thrombosis (1.3% vs 0.2%, p = 0.01). Hospitalization was significantly shorter in abciximab-treated patients (median 3.1 vs 3.5 days, p <0.001), but total in-hospital costs did not differ significantly (13,413 +/- 5,309 US dollars vs 13,000 +/- 6,006 US dollars, p = 0.13). Rates of the composite end point did not differ significantly during the week after discharge (0.8% vs 0.2%, p = 0.10), nor did component event rates. Abciximab during primary PCI is associated with fewer early adverse outcomes, likely contributing to offset its cost. Hospitalizations after primary PCI are so short, however, that efforts to accelerate discharge with abciximab appear unfeasible, and overall costs remain unchanged.