The prevalence of troponin elevations in patients with cardiac arrest (CA) using newer generation troponin assays when the ninety-ninth percentile is used has not been well described. We studied patients admitted with CA without ST elevation myocardial infarction (MI). Treatment included a multidisciplinary protocol that included routine use of hypothermia for appropriate patients. Serial assessment of cardiac biomarkers, including troponin I was obtained over the initial 24 to 36 hours. Patients were classified into 1 of 5 groups on the basis of multiples of the ninety-ninth percentile (upper reference limit [URL]), using the peak troponin I value: <1×, 1 to 3×, 3 to 5×, 5 to 10×, and >10×. Serial changes between the initial and second troponin I values were also assessed. A total of 165 patients with CA (mean age 58 ± 16, 67% men) were included. Troponin I was detectable in all but 2 patients (99%); all others had peak troponin I values that were greater than or equal to the URL. Most patients had peak troponin I values >10× URL, including patients with ventricular fibrillation or ventricular tachycardia (85%), asystole (50%), and pulseless electrical activity (59%). Serial changes in troponin I were present in almost all patients: ≥20% change in 162 (98%), ≥30% change in 159 (96%), and an absolute increase of ≥0.02 ng/ml in 85% of patients. In conclusion, almost all patients with CA who survived to admission had detectable troponin I, most of whom met biomarker guideline criteria for MI. Given the high mortality of these patients, these data have important implications for MI mortality reporting at CA treatment centers. © 2013 Elsevier Inc. All rights reserved.