Forty-seven patients with mild hypertension and 48 normotensive patients entered a blinded, parallel study in which they received a placebo, 10 mmol/day calcium carbonate (CaCO3), or 20 mmol/day CaCO3. There were no significant differences in blood pressure changes among the groups. In the hypertensive group and in patients with the highest blood pressure there were individual falls in systolic pressure, particularly in the group receiving 10 mmol daily CaCO3. In the hypertensive group the changes were: with placebo, -3 +/- 2/-2 +/- 2 mm Hg; with CaCO3 (10 mmol), -7 +/- 3/-2 +/- 2 mm Hg; and with CaCO3 (20 mmol), -2 +/- 3/1 +/- 2 mm Hg. No change was significant, and no pressure changes of patients taking CaCO3 differed significantly from changes of patients taking placebo. Ten of 33 patients taking placebo, 11 of 31 taking 10 mmol/day CaCO3, and nine of 31 taking 20 mmol/day CaCO3 were classified as responders from their systolic blood pressure fall. These response rates did not differ. Eight patients had falls of systolic blood pressure greater than 15 mm Hg. Five were on 10 mmol/day CaCO3 and three on 20 mmol/day CaCO3. This response was significantly different from that with placebo. Univariate analyses failed to reveal any predictive dietary or biochemical parameter. After 3 months of not taking CaCO3, 12 patients classified as responders, including six of the eight with a fall of 15 mm Hg or more, were rerandomized to placebo or to 20 mmol/day CaCO3. In the rechallenge, responses to CaCO3 and placebo were similar, neither causing a significant pressure fall. Calcium carbonate did not reduce blood pressure. The apparent response in a few patients was not verified by rechallenge. The present study does not support calcium supplementation as a useful nonpharmacological measure for reducing elevated blood pressure.