Diabetes may develop in nondiabetic hypertensive subjects during treatment, but the long-term cardiovascular implications of this phenomenon are not clear. We determined the prognostic value of new diabetes in hypertensive subjects. In a long-term cohort study, 795 initially untreated hypertensive subjects, 6.5% of whom with type 2 diabetes, underwent diagnostic procedures including 24-hour ambulatory blood pressure (BP) monitoring and electrocardiography (ECG). Procedures were repeated after a median of 3.1 years in the absence of cardiovascular events. Follow-up duration was 1 to 16 years (median 6.0). New diabetes occurred in 5.8% of subjects initially without diabetes. Antihypertensive treatment included a diuretic in 53.5% of these subjects, versus 30.4% of those in whom diabetes did not develop (P=0.002). Plasma glucose at entry (P=0.0001) and diuretic treatment on follow-up (P=0.004) were independent predictors of new diabetes. Subsequent to the follow-up visit, a first cardiovascular event occurred in 63 subjects. Event rate in nondiabetic subjects at both visits, new diabetes, and diabetes at entry were 0.97, 3.90, and 4.70x100 person-years, respectively (P=0.0001). After adjustment for several confounders, including 24-hour ambulatory BP, the relative risk of events was 2.92 (95% CI: 1.33 to 6.41; P=0.007) in the group with new diabetes and 3.57 (95% CI: 1.65 to 7.73; P=0.001) in the group with previous diabetes, when compared with the group persistently free of diabetes. In treated hypertensive subjects, occurrence of new diabetes portends a risk for subsequent cardiovascular disease that is not dissimilar from that of previously known diabetes.