Abciximab and eptifibatide have been shown to reduce ischemic complications compared with heparin alone in patients with acute coronary syndromes who undergo percutaneous coronary intervention. Whether 1 agent is safer and/or more effective has not been prospectively examined. The aim of this study was to assess the outcomes related to downstream glycoprotein IIb/IIIa inhibitor treatment selection during percutaneous coronary intervention in 2,211 patients with moderate and high-risk acute coronary syndromes in the prospective multicenter Acute Catheterization and Urgent Intervention Triage Strategy (ACUITY) trial. The protocol permitted operator selection of abciximab (n = 835) or eptifibatide (n = 1,376) for routine use in the trial. Multivariate and propensity-based adjustments were used to assess the independent association of glycoprotein IIb/IIIa inhibitor treatment selection with prespecified study end points. Compared to patients receiving eptifibatide, those administered abciximab were older, more likely to be enrolled outside of North America, more frequently had biomarker elevations and ST-segment deviation, but had fewer baseline cardiac risk factors and previous revascularization procedures. After multivariate propensity-based adjustment, abciximab was independently associated with significantly fewer net clinical adverse events (odds ratio 0.61, 95% confidence interval 0.42 to 0.90, p = 0.01), mediated by composite ischemia (odds ratio 0.61, 95% confidence interval 0.38 to 0.98, p = 0.04) and major bleeding (odds ratio 0.58, 95% confidence interval 0.34 to 1.00, p = 0.051). In conclusion, in this prespecified but nonrandomized comparison in patients with acute coronary syndromes who underwent percutaneous coronary intervention with catheterization laboratory initiation of glycoprotein IIb/IIIa inhibitors, the use of abciximab rather than eptifibatide was associated with improved clinical outcomes at 30 days. These findings should be viewed as exploratory in light of the nonrandomized and heterogeneous nature of the comparator groups and significant potential for residual confounding.