Chronic acid loads are an obligate consequence of the high animal/grain protein content of the Western diet. The effect of this diet-induced metabolic acidosis on bone mass is controversial. In a randomized, prospective, controlled, double-blind trial, 161 postmenopausal women (age 58.6 +/- 4.8 yr) with low bone mass (T score -1 to -4) were randomly assigned to 30 mEq of oral potassium (K) citrate (Kcitrate) or 30 mEq of K chloride (KCl) daily. The primary end point was the intergroup difference in mean percentage change in bone mineral density (BMD) at lumbar spine (L2 through L4) after 12 mo. Compared with the women who received KCl, women who received Kcitrate exhibited an intergroup increase in BMD (+/-SE) of 1.87 +/- 0.50% at L2 through L4 (P < 0.001), of 1.39 +/- 0.48% (P < 0.001) at femoral neck, and of 1.98 +/- 0.51% (P < 0.001) at total hip. Significant secondary end point intragroup changes also were found: Kcitrate increased L2 through L4 BMD significantly from baseline at months 3, 9, and 12 and reached a month 12 increase of 0.89 +/- 0.30% (P < 0.05), whereas the KCl arm showed a decreased L2 through L4 BMD by -0.98 +/- 0.38% (P < 0.05), significant only at month 12. Intergroup differences for distal radius and total body were NS. The Kcitrate-treated group demonstrated a sustained and significant reduction in urinary calcium excretion and a significant increase in urinary citrate excretion, with increased citrate excretion indicative of sustained systemic alkalization. Urinary bone resorption marker excretion rates were significantly reduced by Kcitrate, and for deoxypyridinoline, the intergroup difference was significant. Urinary net acid excretion correlated inversely and significantly with the change in BMD in a subset of patients. Large and significant reductions in BP were observed for both K supplements during the entire 12 mo. Bone mass can be increased significantly in postmenopausal women with osteopenia by increasing their daily alkali intake as citrate, and the effect is independent of reported skeletal effects of K.