1. Twenty-six patients under chronic hemodialysis for p a s t two years, 4.1968 to 3.1970, . were divided into the outpatient and inpatient groups and further subdivided into the complete and incomplete groups based on the clinical findings and social adaptability. Findings such as blood pressure, cardiothoracic ratio, urine volume, creatinine clearance were all better in the outpatient group than in the inpatient. 2. For hemodialysis on the o u tpatient basis or on rehabilitation to be possible, the clinical indices should be as follows : mean blood pressure below 100 mmHg, cardiothoracic ratio below50 %, urine volume over 1000 ml per day and creatinine clearance over 4.0 ml/min. Social rehabilitation would be greatly depending on how much is the r e maining function, of the patients' impaired kidneys. In most of the western counteries, chronic hemodialysis is carried out on the outpatient basis. Hemodialysis under hospitalization is, of course, not what we willingly do. If a patient could not be discharged after two months, he should rather have renal transplantation than continuing hemodialysis. 3. Seven patients whose urine vo l ume is over 1000 m l/day were chosen for study of renal function. Changes of renal function were checked for three days among which 8 hours hemodialysis was inserted. After dialysis, GFR and free water clearance dropped, and all of the urnie volume, solute excretion, urinary sodium and potassium decreased. Values per each nephron also decreased. Specific gravity and osmolality showed no change due to hemodialysis and stayed fixed at the low values. So it was conjectured that function of the patients' own kidneys dropped during hemodialysis. After dialysis, GFR, urine volume, solute excretion, urinary sodium and potassium all increased gradually recovering to the predialysis values except for free water clearance which increased rapidly. 4. Discussions were made on the concentrating and diluting power of the kidneys, with emphasis on intact nephron hypothesis of Bricker et al. Their theory is not always applicable for dialysis patients, and there seem to be a certain balance among kidney function, dialysis frequency and uremic symptoms. We should not neglect the remaining function of the patients' own kidneys. If dialysis is continued based only on the uremic symptoms, urine volume might decrease. The renal function seems to be autoregulated by the physical demand of the patients. As to solute excretion, dialysis entirely takes place of the kidney ; therefore, it is natural that the excretion by the own kidney may decrease. This might be understood as regulatory adaptive phenomena by Bricker. Increased free water excretion from the remaining nephrons in the immediate postdialysis period is a good evidence of sufficient diluting power as well as of dominant of tubular diluting function over glomerular function. In this postdialysis condition, no glomerulo-tubular balance may exist. 5. If we regard, however, above changes of re n al function during and after hemodialysis as a series of homeostasis occurring in the patients, dialysis is a substitute of glomerular function although it is an unphysiologic, artificial and physical way of filtration and excretion of solutes. Tubular function which ceased during dialysis becomes dominant after that. Thus, glomerulo-tubular balance may be well maintained after all.