BACKGROUND: The degree to which elevated creatine kinase (CK)-MB in the presence of normal CK is predictive of outcome is not well understood despite having been studied for decades. This analysis examined whether normal CK with elevated CK-MB in patients with non-ST-segment elevation acute coronary syndrome (NSTE ACS) is an independent predictor of worse outcomes. A concomitant goal was to contribute insight to the debate over how patients with NSTE ACS should be managed. METHODS: Data for 25,960 patients from the GUSTO IIb, PARAGON A and B, and PURSUIT trials were analyzed. Of these patients, 6402 were excluded from primary analysis because of missing (unmeasured) biomarkers. Patients with complete laboratory data (n = 19,558) were grouped by CK and CK-MB results. To confirm the primary analysis results, data from patients with missing biomarkers were used in an imputation model. RESULTS: Patients were categorized in 1 of 4 groups: normal CK + normal CK-MB; normal CK + elevated CK-MB; elevated CK + normal CK-MB; or elevated CK + elevated CK-MB. For the primary outcome, 180-day death, or myocardial infarction, Kaplan-Meier estimates were 14.9%, 20.8%, 14.5%, and 18.2%, respectively. Regardless of total CK, elevated CK-MB was associated with a 25% to 49% increased relative risk of worse outcomes. Findings from the analyses were verified by the multivariable model. CONCLUSIONS: CK-MB remains a reliable marker for myocardial necrosis and a strong predictor of worse prognosis. All patients with ACS should have CK-MB measurement to search for cardiac ischemia. Patients with elevated CK-MB should receive aggressive management commensurate with their increased risks.