Since its experimental introduction in 1960, hemodialysis has become a widely performed and relatively safe procedure. Therapeutic strategies have been developed, and the numbers of long-term survivors of hemodialysis therapy have been increasing. Hemodialysis therapy was introduced at Sangenjaya Hospital in October 1970, and the 16 patients who have survived for more than 30 years on hemodialysis therapy since its introduction at the hospital were enrolled in this study to investigate the characteristics of long-term hemodialysis patients. For comparison, 50 patients on hemodialysis for less than 30 years were also studied (21 patients with <10 years hemodialysis, 13 with 10-20 years hemodialysis and 16 with 20-30 years hemodialysis). Background information (age, gender, and cause of renal disease), dialysis dose (single pool [sp.] Kt/V), mineral metabolism (serum phosphate), anemia management (serum hemoglobin), and nutrition (serum albumin and reduced interdialytic weight gain) were assessed. Hemodialysis was instituted at 28.7 +/- 6.4 years of age. The primary cause of end-stage renal disease was chronic glomerulonephritis in all of the patients except one, and in that patient it was polycystic kidney disease. As an index of the dialysis dose, sp. Kt/V was 1.2 +/- 0.11. As an index of mineral metabolism, serum phosphate was 5.4 +/- 0.9 mg/dL. As an index of anemia management, serum hemoglobin was 10.2 +/- 1.2 g/dL. As indexes of nutrition, serum albumin was 4.0 +/- 0.2 g/dL and interdialytic weight gain was 4.43 +/- 1.36%. The sp. Kt/V-value, serum phosphate, serum hemoglobin and interdialytic weight gain did not differ between the four different hemodialysis duration groups. Serum albumin was lower in the >30 group (4.0 +/- 0.2 g/dL) than in the <10 group (4.2 +/- 0.3 g/dL) (P = 0.046). As the duration of hemodialysis has increased, the age at hemodialysis induction has become younger. The cause of the renal failure was chronic glomerulonephritis in most of the cases. None had diabetic nephropathy. Improvement of the prognosis of patients with diabetic nephropathy is required. Most of the indexes of these patients nearly satisfied the recommended values.