The aims of this study were to assess the impact of diabetes and associated variables (fasting plasma glucose, blood pressure, antidiabetic treatment, body mass index) on general and cause-specific mortality in an Italian population-based cohort with Type II (non-insulin-dependent) diabetes mellitus, comprising mainly elderly patients. The patients (n = 1967) who had Type II diabetes were identified in 1988 with an 80% estimated completeness of ascertainment. In 1995, a mortality follow-up (98% completeness) of the cohort was done amounting to a total of 11153 person-years. Observed and expected number of deaths were 577 and 428.7, respectively, giving a standardized mortality ratio (SMR) of 1.35 (95% CI 1.24-1.46). The most common underlying causes of death were malignant neoplasm, ischaemic heart disease and cerebrovascular diseases, which accounted for 18%, 17.8% and 17.5% of deaths, respectively. Cardiovascular disease as a whole (international classification of disease ICD-9 390-459) accounted for 260 of 577 deaths (SMR 1.21, 95% CI 1.07-1.36). In internal analysis, the most important predictors of general mortality were insulin-treatment (relative risk [RR] 1.72, 95% CI 1.19-2.49) and a fasting plasma glucose greater than 8.89 mmol/l ([RR] 1.29, 95 % CI 1.04-1.60), whereas the most important predictors of cardiovascular diseases were insulin-treatment and hypertension. In conclusion, this population-based study showed: 1) slight mortality excess of 35% in Type II diabetes being associated with 2) a 30% increased mortality in subjects with baseline fasting glucose greater than 8.89 mmol/l and 3) a 40% increased risk of death from cardiovascular diseases in hypertensive patients.