OBJECTIVE To evaluate the correlation between predialysis glycemic control and clinical outcomes for type II diabetic patients on continuous ambulatory peritoneal dialysis (CAPD). METHODS Sixty type II diabetic patients on CAPD were classified into 2 groups according to the status of glycemic control. In group G (good glycemic control), more than 50% of blood glucose determinations were within 3.3-11 mmol/L and the glycosylated hemoglobin (HbA1C) level was within 5-10% at all times. In group P (poor glycemic control), fewer than 50% of blood glucose determinations were within 3.3-11 mmol/L or HbA1C level was above 10% at least once during the follow-up duration. In addition to glycemic control status, predialysis serum albumin, cholesterol levels, residual renal function, peritoneal membrane function, and the modes of glycemic control were also recorded. METHODS Dialysis Unit, Department of Nephrology of a single university hospital. METHODS From February 1988 to October 1995, 60 type II diabetic patients receiving CAPD for at least 3 months were enrolled. METHODS Morbidities before and during the dialysis period, patient survival, and causes of mortality. RESULTS The patients with good glycemic control had significantly better survival than patients with poor glycemic control (p < 0.01). There was no significant difference in predialysis morbidity between the two groups. No significant differences were observed in patient survival between the patients with serum albumin greater than 30 g/L and those with less than 30 g/L (p = 0.77), with cholesterol levels greater or less than 5.18 mmol/L (p = 0.73), and with different peritoneal membrane solute transport characteristics evaluated by peritoneal equilibration test (p = 0.12). Furthermore, there was no significant difference in survival whether the patients controlled blood sugar by diet or with insulin (p = 0.33). Cardiovascular disease and infection were the major causes of death in both groups. Although good glycemic control predicts better survival, it does not change the pattern of mortality in diabetics maintained on CAPD. CONCLUSIONS Glycemic control before starting dialysis is a predictor of survival for type II diabetics on CAPD. Patients with poor glycemic control predialysis are associated with increased morbidity and shortened survival.