To evaluate potassium(K) homeostasis during in-terdialytic and dialytic phases in chronic hemodialysis patients, we analyzed pre- and post- dialysis plasma K concentration(n=28) over n week with an interdialytic interval of 7Zhrs, 48hrs(l), and 48hrs(II), respectively, and the quantity of total dialytic K removal via dialysate. The predialysis plasma K at 72h interval(prePK72h: 4.89+/-0.17mEq/L) was significantly higher than those at 48h interval(prePK48h-I: 4.57+/-0.15mEq/L, and prePK48h-II: 4.40+/-15mEq/L) (p=0.000, p=0.000). 10.7% in prePK72h were categorized into severe hyperkalemia more than 6.0mEq/L, but none in prePK48h-I, II(p=0.000, p=0.000). In contrast no difference between 72-h and 42-h intervals was found in the postdialysis plasma K(postPK72h: 3.59+/-0.07 vs postPK48h-I : 3.530+/-08mEq/L, p>0.05) and in the quantity of total dialytic K removal via dialysate(delta Ktota172h : 74+/-2.6 vs delta Ktota148h-I:71+/-2.2mEq, p>0.05). On approach to this with two-compartment model, there was significant difference in dialytic K removal from ECF(delta Kecf72h:22.2+/-1.6 vs delta Kecf48h-I:17.7+/-1.6mEq, p<0.01), but not in that from ICF(delta Kicf72h:51.6+/-3.1 vs delta Kicf48h-I: 53.5+/-2.7mEq, p>0.05). In all 28 patients, age, sex and body weight were not correlated with either pre- and post- plasma K levels or total K removal per kg body weight. In conclusion, the majority of dialytic K removal is from the replenishment of the ICF potassium and it has rather constant feature in that there was no autoregulatory increment even with the higher predialysis plasma K concentration. So the plasma K concentration on chronic maintenace hemodialysis is more dependent on the potassium gain during interdialytic phase than the potassium removal during dialytic phase. Also it is reasonable to restrict dietary K intake and apply K-exalate orientating to the interdialytic phase of 72hrs because severe hyperkalemia is rare in that of 48hrs.