Atherosclerotic vascular disease is a major cause of morbidity and mortality in insulin-dependent diabetes mellitus. The frequent coexistence in these patients of microangiopathy and coronary artery disease was observed more than 30 years ago and later verified in large epidemiological studies. Thus, the subgroup (30-40%) of patients who develop clinical nephropathy, also are at extremely high risk of early cardiovascular death. A number of established cardiovascular risk factors are present not only in advanced clinical nephropathy but also in its earliest stages. These include elevated blood pressure, atherogenic changes in the plasma concentrations of lipids and lipoproteins, elevated plasma levels of fibrinogen and probably hyperreactivity of platelets. However, it seems unlikely that these risk factors fully explain the excess cardiovascular morbidity and mortality in insulin-dependent diabetic patients with clinical nephropathy. Patients with slightly elevated urinary albumin excretion are at increased risk of developing not only clinical nephropathy and coronary heart disease but also proliferative retinopathy and cardiomyopathy. We have, therefore, hypothesised that elevated urinary albumin excretion is a marker of generalized disease in the vascular wall of small and large blood vessels. Findings of elevated transcapillary escape rate of albumin, elevated plasma concentration of von Willebrand factor and impaired fibrinolytic capacity in early diabetic nephropathy have supported this hypothesis. However, the initial pathophysiological mechanisms involved are still hypothetical and largely unknown. During recent years the incidence of clinical nephropathy has declined and the prognosis of insulin-dependent diabetic patients has improved. Whether intervention directed against the often clustered cardiovascular risk factors will further improve the prognosis in proteinuric patients is suggested but still unknown. However, the key question is still, why is the vascular wall, in small and large blood vessels, vulnerable in some but not all diabetic patients? In the future more studies of the initial pathophysiological mechanisms involved in this vulnerability are needed.