International and geographical differences in the survival rates of chronic dialysis patients can be explained by differences in primary renal disease, in the acceptance rate of elderly patients, and in predialysis comorbid conditions. Several studies have shown the effects of these factors on survival. However, in most studies, a large number of patients may leave for renal transplantation or transfer to other centers, so that precise analysis becomes impossible. Although the number of patients in our registry is not so large (n = 1,982), we have few such problems and were able to examine the effects of the above-mentioned factors on patient survival using the Cox proportional hazard model. Hazard ratios (HR) and 95% confidence intervals were 0.739 and 0.366-1.491 in patients with polycystic kidney disease (n = 38), 2.669 and 1.513-4.708 in patients with systemic lupus erythematosus (n = 39), 1.245 and 0.935-1.660 in patients with nephrosclerosis (n = 122), 1.815 and 1.447-2.229 in patients with diabetes mellitus (n = 374), and 1.595 and 1.201-2.117, respectively, in patients with other renal diseases (n = 146) when the HR in patients with chronic glomerulonephritis (n = 1,263) was taken as 1.00. HR and 95% confidence intervals were 1.222 and 1.016-1.470 in patients with one comorbid condition (n = 217) and 1.494 and 1.033-2.160, respectively, in patients with two comorbid conditions (n = 24) when the HR of patients with no predialysis comorbid conditions (n = 1,741) was taken as 1.00. Our data demonstrate the effects of renal diseases and number of predialysis comorbid conditions on the survival in chronic dialysis patients. Differences in proportion of diabetic patients and age at entry may cause large differences in survival rates. In diabetic patients, causes of malnutrition and the effects of dialysis dose on the survival rate remain to determined.