Small doses of (NH4)2HPO4 or KH2PO4 by stomach tube caused increase in plasma P04 and P04 excretion. Above a threshold of 0*8 mmol. 1-1, increase of plasma P04 by 0 5 mmol. l-1 caused P04 excretion to increase by about 35 ,umol. min.-' After KH2PO4 this relationship was not altered by the concurrent increases in plasma K and K excretion. After doses of (NH4)2SO4 or K2SO4, excretion of SO4 was similarly related to plasma S04 and was independent of plasma K and K excretion.An effect of P04 on K excretion was observed after doses of (NH4)2HP04, when increased excretion of P04 was accompanied by increased excretion of K without change in plasma K. There was also increased excretion of NH4 and a small increase in Na excretion. The changes were similar to those produced by (NH4)2SO4 [O'Connor and Summerill, 1976].KH2PO4 and K2SO4 produced increase in plasma K and increased excretion of K not significantly different from the changes produced by KCI or KHCO3 [Baylis and O'Connor, 1976]. After KH2PO2 or K2SO4, the urinary anion was P04 or SO4, instead of Cl and HCO3. Any effect of anions on K excretion was much less than the effect of increase in plasma K.At low rates of excretion of K, increased urinary excretion of impermeant anion can determine increased excretion of K. However, the effect of anion is small in comparison with the effect of increase in plasma K.In a previous paper, Baylis and O'Connor [1976] described the changes in blood and urine produced by small doses of KCI or KHCO3 given to normal conscious dogs, and reached the conclusion that, with K excretion close to normal rates, plasma K is an important determinant of K excretion. This paper is concerned with the possible effect of urinary anions in determining excretion of K. In the experiments of Baylis and O'Connor [1976], K in the urine was accompanied by Cl and HCO3. However, SO4 and P04 normally constitute 70 % of the anion in the urine of dogs. For both SO4 and P04 there is proximal tubular reabsorption and the amount appearing in the urine is the excess of filtered load over maximum tubular reabsorption [Mudge, Berndt and Valtin, 1973]. It is believed that S404 and P04 [Staum, Hamburger and Goldberg, 1972] cannot be reabsorbed in the distal tubule whereas Cl and HCO3 can. Several authors have indicated that K excretion may be increased by impermeant anions such as ferricyanide and sulphate [Berliner, Kennedy and Hilton, 1950;Cort and McCance, 1954;Sullivan, Wilde and Malvin, 1960] but they had used large doses of the salts to produce rates of excretion far greater than occur normally. The purpose of this paper is to study the effect of SO4 and P04 on K excretion close to the normal rates of excretion of the anions and of K. The changes in plasma and urine produced by small doses of (NH 4)2 HPO4, KH2PO4 and K2SO4 have been recorded and the excretion of K compared with that observed by Baylis and O'Connor [1976] after doses of KCl