A crossover design field trial was conducted to investigate the effect of feeding a source of rumen soluble Ca (30%) and Mg (6%) from calcareous marine algae (CMA) on serum and urinary concentrations of Ca and Mg compared with those achieved from finely ground limestone and magnesium oxide (CONT). Periparturient cows (Holsteins and Jerseys) synchronized for seasonal breeding regimen were used (n = 67 control and n = 76 test) in the trial. Diets were formulated to provide 30 g/d of supplemental Ca (17% of total) and 6 g/d supplemental Mg (10% of total) from either CONT or CMA for a minimum of 10 d before calving. Cows that did not consume diets for a minimum of 10 d before calving were not enrolled in the trial. Urine samples were collected daily before calving (average 2.6 d) for analysis of Mg, Ca, and creatinine, which was used as a marker for calculating urine volume based on 29 mg creatinine excreted per kg BW. Blood was collected at 12 h post calving and analyzed for serum total and ionized Ca and Mg. Blood Ca and Mg values were affected by both breed and lactation number, thus breed and lactation were included in the final statistical model as covariates to evaluate the effect of treatment. Cows fed CMA had increased serum concentrations of total Ca (P = 0.027) and ionized Mg (P = 0.020), but no differences were observed in ionized Ca or total Mg (P > 0.10). Urinary concentrations of Ca and Mg were higher (P = 0.01) for cows fed CMA which results in increased (P < 0.0001) excretion of Mg implying that metabolic Mg tissue pools have been saturated. No differences (P > 0.10) were observed in Ca excretion: 12.48 and 12.95 g/d for CONT and CMA, respectively. Regression analysis predicted serum Mg using urine Mg excretion as follows: urine Mg = −0.34 + 2.5 Serum Mg (P < 0.023, R² = 0.08). No differences were observed for other metabolites. These results suggest Mg bioavailability was greater for CMA relative to conventional MgO, however, no improvements in serum Ca were observed.