Background-The relationship between admission glucose levels and outcomes in older diabetic and nondiabetic patients with acute myocardial infarction is not well defined. Methods and Results-We evaluated a national sample of elderly patients (nϭ141 680) hospitalized with acute myocardial infarction from 1994 to 1996. Admission glucose was analyzed as a categorical (Յ110, Ͼ110 to 140, Ͼ140 to 170, Ͼ170 to 240, Ͼ240 mg/dL) and continuous variable for its association with mortality in patients with and without recognized diabetes. A substantial proportion of hyperglycemic patients (eg, 26% of those with glucose Ͼ240 mg/dL) did not have recognized diabetes. Fewer hyperglycemic patients without known diabetes received insulin during hospitalization than diabetics with similar glucose levels (eg, glucose Ͼ240 mg/dL, 22% versus 73%; PϽ0.001). Higher glucose levels were associated with greater risk of 30-day mortality in patients without known diabetes (for glucose range from Յ110 to Ͼ240 mg/dL, 10% to 39%) compared with diabetics (range, 16% to 24%; P for interaction Ͻ0.001).After multivariable adjustment, higher glucose levels continued to be associated with a graded increase in 30-day mortality in patients without known diabetes (referent, glucose Յ110 mg/dL; range from glucose Ͼ110 to 140 mg/dL: hazard ratio [HR], 1.17; 95% CI, 1.11 to 1.24; to glucose Ͼ240 mg/dL: HR, 1.87; 95% CI, 1.75 to 2.00). In contrast, among diabetic patients, greater mortality risk was observed only in those with glucose Ͼ240 mg/dL (HR, 1.32; 95% CI, 1.17 to 1.50 versus glucose Յ110 mg/dL; P for interaction Ͻ0.001). One-year mortality results were similar. Conclusions-Elevated glucose is common, rarely treated, and associated with increased mortality risk in elderly acute myocardial infarction patients, particularly those without recognized diabetes. (Circulation. 2005;111:3078-3086.)