We investigated the rate of decline in GFR and the changing prevalences of micro- and macrovascular complications in 20 type II diabetic patients [mean age 58 (46-71) years, female:male = 7:13, duration of diabetes 16 (12-30) years] from the stage of macroproteinuria with GFRs which were still normal until the beginning of dialysis or the time of death. Controls of renal function, proteinuria, HbAlc, serum lipids, and blood pressure were performed every 6 months at the beginning of the study and later on at 3-month intervals. Fundoscopy, electrocardiogram at rest and in case of need a symptom-limited treadmill ECG, a Duplex ultrasound examination of the carotid vessels, and a Doppler sonographic examination of the femoral arteries were repeated each year. The creatinine clearance (mean +/- SD) of the patients was 81 +/- 6 mL/min/1.73 m2 at the beginning of the study. The rate of decline in creatinine clearance was 1.01 +/- 0.38 mL/min/month during the whole period of observation. Twelve patients (group A) required dialysis after a mean time of 74 (40-119) months; their creatinine clearance was 7 +/- 2 mL/min/month at the beginning of renal replacement therapy. Eight patients (group B) died a short time before the beginning of dialysis treatment; their creatinine clearance was 13 +/- 5 mL/min/1.73 m2. The causes of death were sudden death (n = 4), cardiac failure (n = 1), and stroke (n = 2); in one case it was unknown. The two patient groups did not differ in respect to the mean age, duration of diabetes, HbAlc values, serum cholesterol levels, and blood pressure. The decline in the creatinine clearance was also similar in both patient groups, with 1.07 +/- 0.35 versus 0.98 +/- 0.41 mL/min/month. Only the mean serum triglyceride concentration was significantly higher in the patients who died before dialysis. At the start of the study, cerebrovascular disturbances (including plaques in the carotid vessels) were found in 30%, cardiovascular disturbances (including pathologic ECG findings) in 45%, a peripheral vascular disease in 15%, and diabetic retinopathy (grade I and II) in 75%. At the beginning of dialysis treatment or the time of death, respectively, the prevalence of cerebrovascular diseases was increased to 70% and the prevalence of cardiovascular diseases to 90%; peripheral vascular disease was present in 50% and diabetic retinopathy in all of the cases. We conclude that type II diabetic patients show high mortality (40%) and poor quality of life, not only when they require dialysis treatment, but also in the predialysis phase.