A study was carried out to verify the effect of Ca and P levels on production, digestibility, and serum bone metabolism biomarkers in dairy cows. Fifty-two nonlactating multiparous cows (>or=3 lactations) were confined in a free-stall barn approximately 20 d before calving. A standard close-up diet was fed to cows once daily until d 2 postpartum. Cows were randomly assigned to 1 of 4 dietary treatments arranged in a 2 x 2 factorial approach averaging 0.64% Ca for high Ca (HCa), 0.46% Ca for low Ca (LCa), 0.47% P for high P (HP), and 0.38% P for low P (LP) on a dry matter basis. Experimental diets were fed twice daily from 3 d in milk (DIM) until 31 DIM. Intake and milk yield were recorded daily. Milk samples were collected on d 28, 29, and 30 postpartum for components analyses. Blood samples were drawn 10 d before expected calving, at calving, and at 15 and 30 DIM for serum analyses of osteocalcin, a biomarker of bone accretion, and pyridinoline, a biomarker of bone resorption. Total fecal collection was conducted when cows in a block averaged 20 DIM. Intake and production traits were not significantly affected by any of the dietary treatments. Cows averaged nearly 21 kg/d dry matter intake and 44 kg/d milk yield from 6 to 31 DIM. There were no significant differences across treatments in body weight or body condition score loss. Phosphorus intake, P fecal output, P digestibility, and P apparent absorption were affected by dietary P content. Calcium intake was higher with HCa, but Ca fecal output, digestibility, and apparent absorption showed an interaction between dietary Ca and dietary P. Calcium fecal output was 100.6 g/d for cows fed HCaHP, intermediate for cows on the HCaLP diet (89 g/d), and similar among cows fed the 2 LCa diets (70 g/d with LCaHP and 75 with LCaLP). There was no significant effect of Ca or P on osteocalcin measurements. Pyridinoline concentrations were affected by dietary Ca levels and tended to have a significant dietary Ca x dietary P interaction. Phosphorus apparent digestibility occurred independently of dietary Ca levels. Results of this study suggest that more bone was mobilized in cows fed LCa diets, but excess dietary P caused greater and prolonged bone mobilization regardless of dietary Ca content.