Six adult patients (4 females and 2 males, age range 26-57 years) with Gitelman's syndrome (GS) were treated with spironolactone 200-300 mg/day (n = 5) and/or amiloride 10-30 mg/day (n = 3) for 1-18 months. The patients had hypokalemia, hyperreninemia, chloride-resistant metabolic alkalosis, renal hypomagnesemia (n = 5), and hypocalciuria (n = 5). Free water clearance studies during maximal water diuresis and furosemide administration were suggestive of a solute reabsorptive defect beyond the loop of Henle. Antialdosterone therapy induced a significant increase of P-K (from 2.6 +/- 0.4 to 3.4 +/- 0.4 mM; p < 0.0001) and a decrease of C-K (from 21.4 +/- 13.2 to 10.6 +/- 4.8 ml/ min, p < 0.02) and FE(K) (from 21.0 +/- 13.6 to 13.4 +/- 5.7%; p < 0.03); P-Mg increased from 1.38 +/- 0.38 to 1.64 +/- 0.21 mg/dl(p < 0.03) with a parallel fall of C-Mg (from 5.5 +/- 2.3 to 2.9 +/- 1.5 ml/min; p < 0.02) and FE(Mg) (from 5.7 +/- 2.6 to 2.9 +/- 0.6%; p < 0.05); arterial blood pH and HCO3- did not change (P = plasma, C = clearance, FE = fractional excretion). The creatinine clearance fell (from 90.5 +/- 16.8 to 65.8 +/- 20.9 ml/min; p < 0.05), and P-renin rose(from 16.6 +/- 8.9 to 35.3 +/- 25.3 ng/ml/h; p < 0.02, as did P-aldo (from 26.1 +/- 12.3 to 109 +/- 82.6 ng/dl; p < 0.01), indicating extracellular fluid volume contraction; however no significant clinical symptoms of hypovolemia ensued. Despite increased P-aldo levels, estimated transtubular K gradient in K secretory sites fell (from 8.0 +/- 4.0 to 6.7 +/- 3.4; p < 0.01), confirming blunted aldosterone tubular effect. At the dosages employed, spironolactone induced a greater increase of P-K (0.81 +/- 0.52 mM) than amiloride (0.07 +/- 0.41; p < 0.001). In conclusion, antialdosterone therapy is effective in ameliorating hypokalemia and hypomagnesemia in GS. Its effects appear to result mainly from a direct tubular effect on K secretion and Mg reabsorption; extracellular fluid volume contraction appears also to occur during therapy, but has no relevant clinical effects. These results confirm that hypokalemia in GS is more a consequence of increased tubular secretion in the cortical collecting tubule than of impaired tubular K reabsorption; moreover, impaired tubular Mg conservation in GS might also occur in more distal segments of the nephron than previously supposed.