Diabetic patients have substantially greater long-term mortality after acute myocardial infarction (AMI) than nondiabetic patients. Traditionally, cardiac autonomic neuropathy is believed to contribute significantly to the increased mortality rate in patients with diabetes mellitus. In contrast, a recent study suggested that RR-interval variability (RRV) did not predict mortality after AMI in diabetic patients. We compared, in diabetic and nondiabetic patients, the relation between low RRV and long-term mortality in the Multicenter Post Infarction Program (MPIP), a longitudinal observational study of 715 survivors of AMI, including 117 diabetic patients. We studied the association between mortality and 6 frequency-domain measurements and 1 time-domain measurement of RRV. We tested 2 hypotheses: (1) RRV is lower in diabetic patients; and (2) low RRV is less predictive of mortality in diabetic patients. Reduced RRV was significantly more frequent in diabetic patients than nondiabetic patients for all measurements, except high-frequency (HF) power. In diabetic patients, the association between reduced RRV and long-term mortality was at least as strong as it was in nondiabetic patients for all measurements except HF power; this pattern was found for all-cause, cardiac, and arrhythmic death. In multivariable models, reduced RRV was significantly associated with all-cause mortality in diabetic patients even when adjusted for left ventricular ejection fraction, heart failure class, and ventricular arrhythmias. In our post-AMI sample, RRV provided valuable prognostic information among diabetic patients. Our findings suggest that cardiac autonomic neuropathy plays a role in the high mortality rate seen in diabetic patients after AMI.