This study evaluates the use of calcium carbonate in chronic renal failure. Forty-eight patients (25 male, 23 female, mean age 54.3 years, six pre-dialysis, 12 CAPD, 30 haemodialysis) on phosphate restriction and requiring aluminium hydroxide (mean 2.4 +/- 0.8 g/day) to control serum phosphate, were converted to an equivalent dose of calcium carbonate (2.5 +/- 0.6 g/day). None received vitamin D analogues. Three months post-conversion there was a significant decrease in mean (+/- SEM) serum phosphate (1.86 +/- 0.08 versus 1.66 +/- 0.05 mmol/l, P < 0.01) and serum aluminium (28.3 +/- 5.4 versus 13.2 +/- 3.0-mu-g/l, P < 0.0001); calcium/phosphate product was unchanged. Post-conversion there was an increase in serum bicarbonate, (20.6 +/- 0.5 versus 22.1 +/- 0.6 mmol/l, P < 0.01) and serum calcium (2.32 +/- 0.02 versus 2.45 +/- 0.03 mmol/l, P < 0.0001). No change in serum creatinine, alkaline phosphatase or parathormone occurred. No adverse effects were reported but nine (18%) patients became hypercalcaemic (2.7 to 2.93 mmol/l), eight of whom responded to dose reduction. Hypercalcaemia did not correlate with pre-conversion serum calcium, parathyroid hormone, alkaline phosphatase or aluminium. Calcium carbonate is an effective alternative to aluminium-based phosphate binders. It produces a beneficial increase in serum calcium and bicarbonate and a significant decrease in serum aluminium. Hypercalcaemia is unpredictable but is easily reversible in the majority of patients.