Renal dysfunction is an important predictor of mortality in patients with acute coronary syndrome. Until recently, the Cockcroft-Gault (C-G) equation has been most commonly used to estimate renal function, although the Modification of Diet in Renal Disease (MDRD) equation is now recommended. Which equation better predicts mortality is unclear. Consecutive patients without ST elevation on the initial electrocardiogram admitted for exclusion of myocardial ischemia were included. Admission creatinine was used to estimate renal function, and 30-day and 1-year mortality were compared after classifying patients as having no (estimated glomerular filtration rate [eGFR] > or =60 ml/min/m(2)), moderate (eGFR 30 to 59 ml/min/ m(2)), or severe (eGFR <30 ml/min/m(2)) renal dysfunction using the 2 equations. Of the 4,343 patients (49% women, 64% African-American) included, 16% had troponin I elevations consistent with myocardial infarction, and 1-year mortality was 10%. Agreement between the 2 equations was high (r = 0.87 p <0.001, concordance 86%). Mortality was similar in the 2 renal function classifications, with no significant differences based on race or troponin I status at 30 days or 1 year. However, the area of the receiver operator characteristic curve was significantly larger for predicting 1-year mortality with the C-G equation (0.75 [0.72 to 0.77] vs 0.71 [0.68 to 0.73]; p <0.01); both were superior to creatinine alone (0.68 [0.66 to 0.71]; p <0.01 for both C-G and MDRD). Results for 30-day mortality were similar. In conclusion, the C-G equation appears to be superior to the MDRD equation for predicting short- and long-term mortality in patients admitted for exclusion of ischemia, although differences are minor.