We measured the rate of oxalate flux across the red-cell membrane in the steady state in 114 patients with a history of calcium oxalate kidney stones and in 25 controls. Of the patients, 98 had recurrent, "idiopathic" kidney stones, 8 had primary hyperparathyroidism, 7 had renal or urinary tract malformations, and 1 had primary hyperoxaluria. Oxalate exchange was significantly higher in the 98 patients with idiopathic stone formation than in the controls (-1.10 +/- 0.95 [SD] X 10(-2) min-1 vs. -0.31 +/- 0.12 X 10(-2); P less than 0.001); it was above the upper limits of normal in 78 of these patients. All 8 patients with hyperparathyroidism and the patient with primary hyperoxaluria had values in the normal range; 2 of the patients with renal or urinary tract malformation had values at the upper normal limit. A study of five families indicated that the abnormality is an autosomal monogenic dominant trait with complete penetrance and variable expressivity. Oxalate-tolerance tests were carried out in five pairs of brothers. One brother in each pair had the abnormality in oxalate flux, and had a significantly higher percentage of oxalate excretion at two hours after oxalate loading (18.09 +/- 3.07 [SD] vs. 10.37 +/- 3.08 percent; t = 3.97; P less than 0.005) and four hours (14.87 +/- 2.91 vs. 9.89 +/- 2.93 percent; t = 2.70; P less than 0.05). Treatment with oral hydrochlorothiazide (50 mg per day) or amiloride (5 mg per day) or both restored normal or nearly normal red-cell oxalate exchange in all of 33 patients who initially had increased rates. We conclude that an inherited cellular defect in oxalate transport may be a factor in "primary" calcium oxalate stone formation and that this defect may be corrected with diuretics.