To test the hypothesis that interaction between genetic, immunological, clinical and metabolic risk factors influences the outcome of Type II (non-insulin-dependent) diabetes mellitus, we examined which of the above factors present at baseline were associated with mortality in 134 Type II diabetic patients followed for 9 years. Thirty-eight patients (29 %) died during the follow-up period; the majority of whom (68 %) died from cardiovascular disease. At baseline, the deceased patients had higher HbA_{1 c} values (p = 0.002), higher LDL-triglycerides (p = 0.007), lower HDL-cholesterol (p = 0.007), higher non-esterified fatty acid (NEFA) concentrations (p = 0.014), and higher albumin excretion rate (p < 0.0001) than the patients who survived. In addition, the frequency of HLA-DR4 (21 vs 39 %, p = 0.048) and of parietal cell antibodies (5 vs 14 %, p = 0.016) were decreased in the deceased as compared to the living patients. Patients who died during follow-up also had more retinopathy (42 vs 16 %, p = 0.002), neuropathy (57 vs 23 %, p < 0.001), microalbuminuria (45 vs 6 %, p < 0.0001), coronary heart disease (50 vs 13 %, p < 0.0001), and peripheral vascular disease (27 vs 9 %, p = 0.005) at baseline than patients who survived. In a multiple logistic regression analysis macroangiopathy (p = 0.004), neuropathy (p = 0.007), HbA_{1 c} (p = 0.018) and albumin excretion rate (p = 0.016) were independent risk factors for death. In patients free of cardiovascular disease at baseline, conventional risk factors such as LDL-cholesterol (p = 0.005) and age (p = 0.003) were associated with subsequent development of cardiovascular disease. In conclusion, in addition to coexisting macroangiopathy, increased albumin excretion rate, poor glycaemic control and neuropathy are risk factors for cardiovascular mortality in patients with Type II diabetes. The presence of HLA-DR4 and signs of autoimmunity may be associated with decreased risk of cardiovascular disease. [Diabetologia (1998) 41: 1253–1262]