Peritoneal dialysis is generally accepted as a major therapeutic procedure for renal failure, uremia, and distortions in blood electrolytes and other chemical components. Although its dialyzing effect is less efficient than hemodilysis, peritoneal dialysis is employed frequently as appropriate means of treat-ment for these disorders because of technical simplicity and absence of complicated or expensive apparatus. Advanced understanding of water and electrolyte metabolism as well as availability of more potent diuretics provide the more efficient measures in the treatment of edema. However, the treatment of refractory edema, congestive heart failure and pulmonary edema appearing in patients with renal failure are still eluded the adequate control. In such overhydrated states hemodialysis and peritoneal dialysis with hypertonic dialysis fluid may be the treatment of choice. The purpose of this paper is to evaluate peritoneal dialysis comparing its effect with that of hemodialysis, to analyse the kinetics of peritoneal dialysis with hypertonic irrigation fluid, to define the influences on circulation, and finally to elucidate the clinical effectiveness of dialysis with hypertonic fluid for removal of excess fluid. 1) Intermittent peritoneal dialysis were performed on 22 patients in 86 occasions using commercially available peritoneal dialysis fluid. An amelioration of uremic signs and symptomes with improvement in azotemia, electrolyte abnormalities and metabolic acidosis were obtained. Peritoneal dialysis was about one-sixth as efficient as Kolff Twin-Coil artificial kidney. Peritonitis was in 3 cases which were dialyzed with indwelling peritoneal catheters. On the other hand, peritonitis was not observed in cases dialyzed with multipuncture technics. 2) Experiments were performed with mongrel dogs of 8-14 kg. to evaluate the peritoneal dialysis with a dialysis fluid containing 7.0% dextrose and its influence upon circulation. The dogs were divided into three groups ; dehydrated normal dogs (group I, 8 dogs), dehydrated uremic dogs (group II, 4 dogs) and overhydrated uremic dogs (group III, 4 dogs). The dogs were made uremic by ligation of bilateral ureters and used for the experiments between 2 and 4 days after the operation. Serial determinations of intraperitoneal fluid were carried out on the osmolality and the isotope, sodium and chloride concent-rations after the intraperitoneal administration of 500 ml. of hypertonic dialysis fluid containing 7.0% dextrose with 10-20 pc of RISA. Intraperitoneal fluid volumes were calculated from isotope concentra-tions. Predialysis plasma osmolality of group II and III was higher than that of group I. Soon after the initiation of dialysis, the fluid osmolality fell rapidly and leveled off in 120 minutes (group II and III) or 180 minutes (group I) when the fluid osmolality was in equilibrium with plasma osmolality. Intraperitoneal fluid increased rather rapidly in the first 30 minutes and no significant difference was observed among three groups in their time course profil. Thus, the most efficient peritoneal dialysis can be obtained when the dialysate is changed every 30 minutes. A rapid decrease in sodium and chloride concentration, possibly due to a shift of water into the peritoneal cavity, occurred reaching the nadir from 60 to 120 minutes. The ratio of sodium to water lost during this procedure was in favor of water, indicating water being removed in excess of sodium. Blood pressure decreased in group I, stayed unchanged in group II and III. Increase of hematocrit and decrease of plasma volume were far greater in group I than in II and III. There were no significant difference in the intraperitoneal fluid volume among three groups. These facts suggest that less untoward influences upon circulation iu group II and III were probably, at least in part, due to overhydration in these groups. 3) Clinically, about 500 ml. of fluid removal were attained when one liter each