QT-interval parameters are potential indicators of increased cardiovascular risk. We evaluated prospectively their prognostic value, in relation to other risk markers, for cardiovascular fatal and nonfatal events in a cohort of 271 hypertensive type 2 diabetic outpatients. QT intervals were measured from 12-lead standard ECGs obtained on admission and maximum rate-corrected QTinterval duration and QT-interval dispersion (QTd) calculated. Clinical and laboratory data and 2-D echocardiograms (available in 126 patients) were recorded. Survival analyses included Kaplan-Meier survival curves, uni and multivariate Cox proportional-hazards models. After a median follow-up of 55 months (range 2-84), 68 total fatal or nonfatal cardiovascular events and 34 cardiovascular deaths (24 of them from cardiac causes) were observed. In multivariate Cox analysis, QTd was an independent predictor for total cardiovascular events (HR: 1.16, 95% CI: 1.01-1.34, for each 10 ms increments) and for cardiac deaths (HR: 1.28, 95% CI: 1.01-1.60). Other independent risk indicators for cardiovascular morbidity and mortality were echocardiographic left ventricular hypertrophy (Echo-LVH), serum triglycerides, presence of pre-existing cardiac and peripheral arterial disease, age, diabetes duration, heart rate and the presence of frequent ventricular premature contractions on ECG. The combination of QTd and Echo-LVH improved cardiovascular risk stratification compared with either alone, the presence of both prolonged QTd (465 ms) and Echo-LVH was associated with a 3.2-fold (95% CI: 1.7-6.1) increased risk of a first cardiovascular event and a 5.9-fold (95% CI: 2.1-16.4) increased risk of cardiovascular death. Thus, QT provided additive prognostic information for cardiovascular morbidity and mortality beyond that obtained from conventional risk markers, including Echo-LVH, in type 2 diabetic patients with arterial hypertension.