Calcium carbonate has been successfully used as a phosphate binder in patients with chronic renal failure; however, a high frequency of hypercalcaemia has been reported. To study the effects of calcium carbonate preparations with different dissolution characteristics on the incidence of this side effect, we conducted a double-blind, crossover trial in 21 patients undergoing chronic haemodialysis. Aluminum hydroxide therapy was replaced with calcium carbonate. The subjects then randomly received either an enteric-coated or a gastric-coated preparation. Calcium carbonate (3.1-3.6 g/d) controlled serum phosphate concentrations as effectively as aluminium hydroxide (2.9 g/d). Concurrently, there was a significant rise in mean serum calcium and a fall in serum concentrations of both parathyroid hormone and osteocalcin, the latter suggesting a decrease in bone turnover. Overall, hypercalcaemic episodes developed in 9 patients (43%) and occurred at a considerable frequency (33 episodes per 100 patient-months) during treatment with the gastric-coated formulation. Following conversion to enteric-coated calcium carbonate (3.6 g/d) patients had fewer occurrences of hypercalcaemia (12 episodes per 100 patient-months, P less than 0.05) and, as compared to the gastric-coated preparation, increases in serum calcium greater than 3.00 mmol/l were not observed at all. Hyperaluminaemia was regressive during therapy with calcium carbonate, but addition of small doses of aluminium hydroxide caused a large rise in serum aluminium concentrations after infusion of desferrioxamine, indicating an enhanced rate of absorption or aberrant compartmentalization of aluminium. We conclude that calcium carbonate can control hyperphosphataemia in dialysis patients.(ABSTRACT TRUNCATED AT 250 WORDS)