1. Muscle electrolyte (Na, K, Cl, P) concentrations and fluid compartments of six children with Bartter's syndrome were compared with those of 17 control children. From four of the patients biopsies were taken twice, before and after treatment with potassium chloride and spironolactone. Muscle electrolytes were determined by neutron activation in tissue sampled by needle biopsy. Results were expressed in relation to fat-free dry solids. Trapped extracellular fluid was calculated from the chloride space, a normal resting membrane potential being assumed.
2. In hypokalaemic patients total muscle potassium (Km) (33·0 ± 2·1 sem) was below the normal range (43·8 ± 0·7 mmol/100 g of fat-free dry solids) in all cases. When plasma potassium was corrected, total muscle potassium returned to normal. There was a significant correlation between plasma potassium concentration and total muscle potassium (r = 0·87; P < 0·01). Total muscle sodium (Nam) was increased: 18·7 ± 2·9 mmol/100 g in hypokalaemic patients (11·3 ± 0·6 mmol/100 g of fat-free dry solids in control subjects). A negative relationship was found between Km and Nam (r = —0·85; P < 0·01). With treatment, intracellular potassium concentration remained low (142·0 mmol/l of intracellular water; control, 163·0 ± 2·4 mmol/l) despite normokalaemia and normal total muscle potassium. Muscle phosphorus (32·1 ± 1·8 mmol/100 g) was higher than in control subjects (28·1 ± 0·4 mmol/100 g of fat-free dry solids).
3. These data suggest that sodium replaces potassium in muscle of patients with Bartter's syndrome. The low intracellular potassium concentration in the binephrectomized sister of one child and in all treated children suggests that there are generalized anomalies in the handling of cations in Bartter's syndrome.