The prognostic importance of electrocardiographic ventricular repolarization QT parameters (maximum rate-corrected QT interval-QTcmax, QT interval dispersion-QTd, and QTcd), in relation to other risk markers, on cardiovascular and cardiac mortality, and on total fatal or nonfatal cardiovascular events, was evaluated prospectively in 471 type 2 diabetic outpatients. During a median follow-up of 57 months (range: 2-84), 121 (25.7%), patients died, 44 (36.3% of them) from cardiovascular causes and 106 (22.5%) fatal or nonfatal cardiovascular events were observed. In Cox proportional hazards multivariate analysis, both QTd and QTcmax were independent predictors of cardiovascular and cardiac mortality (hazard ratio [HR]: 1.34, 95% confidence interval [95% CI]: 1.12-1.59, for each 10-ms increments in QTd and HR: 1.17, 95% CI: 1.03-1.21 for 10-ms increments in QTcmax, for cardiovascular mortality). They were also predictors of total fatal or nonfatal cardiac and cardiovascular events (HR: 1.18, 95% CI: 1.05-1.33 for QTd and HR: 1.09, 95% CI: 1.04-1.15 for QTcmax). Additional independent prognostic markers for total cardiovascular events were the presence of previous cardiac disease, cerebral or peripheral vascular disease, age, male gender, known diabetes duration, heart rate, and serum triglycerides. Excluding patients with prior cardiac disease did not change significantly the prognostic performance of QTd but decreased that of QTcmax. In conclusion, QT interval parameters give additional prognostic information in patients with type 2 diabetes mellitus, beyond that obtained from traditional risk factors. QT interval dispersion seems a better prognostic marker than maximum QT interval, particularly in patients without previous cardiac diseases.