Baseline electrocardiogram abnormalities and marker elevations not associated with myocardial necrosis make accurate diagnosis of myocardial infarction (MI) difficult in patients with cocaine-associated chest pain. Troponin sampling may offer greater diagnostic utility in these patients. Objective: To assess outcomes based on troponin positivity in patients with cocaine chest pain admitted for exclusion of MI. Methods: Outcomes were examined in patients admitted for possible MI after cocaine use. All patients underwent a rapid rule-in protocol that included serial sampling of creatine kinase (CK), CK-MB, and cardiac troponin I (cTnI) over eight hours. Outcomes included CK-MB MI (CK-MB Ն 8 ng/mL with a relative index [(CK-MB ϫ 100)/total CK] Ն 4, cardiac death, and significant coronary disease (Ն50%). Results: Of the 246 admitted patients, 34 (14%) met CK-MB criteria for MI and 38 (16%) had cTnI elevations. Angiography was performed in 29 of 38 patients who were cTnI-positive, with significant disease present in 25 (86%). Three of the four patients without significant disease who had cTnI elevations met CK-MB criteria for MI, and the other had a peak CK-MB level of 13 ng/mL. Sensitivities, specificities, and positive and negative likelihood ratios for predicting cardiac death or significant disease were high for both CK-MB MI and cTnI and were not significantly different. Conclusions: Most patients with cTnI elevations meet CK-MB criteria for MI, as well as have a high incidence of underlying significant disease. Troponin appears to have an equivalent diagnostic accuracy compared with CK-MB for diagnosing necrosis in patients with cocaine-associated chest pain and suspected MI.