Eighty-two multiparous Holstein cows were enrolled 28 d before expected calving and assigned to 1 of 4 dietary treatments in a randomized block design experiment with a 2 × 2 factorial arrangement of treatments to determine the effect of feeding a neutral or acidogenic diet varying in Ca concentration on prepartum and postpartum intake, blood mineral and metabolite concentrations, and postpartum milk production. Prepartum diets were formulated to provide a dietary cation-anion difference (DCAD) of -21 (negative, NEG) or -2 (neutral, NEU) mEq/100 g of dry matter with either 1.3% or 1.8% Ca. After calving, cows remained on trial through 63 d in milk (DIM) and were fed a common lactation diet. Urine pH was lower for NEG compared with NEU and tended to be lower for 1.8% Ca compared with 1.3% Ca. Fractional excretion of Ca and Mg in urine was greater for NEG than for NEU. Prepartum plasma bicarbonate was lower and P was higher for NEG compared with NEU. Prepartum plasma P and blood urea nitrogen to creatinine ratio was higher for 1.3% compared with 1.8% Ca. Postpartum, concentrations of plasma total protein, albumin, blood urea nitrogen, Mg, and ionized Mg (iMg) were higher and Na was lower for NEU compared with NEG. An interaction of DCAD and Ca was observed for plasma creatinine, which was highest for cows fed NEU and 1.3% Ca compared with all other treatments. Interactions of DCAD and DIM were observed for plasma bicarbonate and iMg. Bicarbonate was higher at 3 DIM and lower at 14 DIM for NEU compared with NEG. Concentrations of iMg were higher at 1, 2, and 14 DIM for NEU compared with NEG. Interactions of Ca and DIM were observed for plasma Ca, Cl, and anion gap. Compared with cows fed 1.5% Ca, those fed 1.3% Ca had lower Ca and anion gap and higher Cl at 1 DIM and lower Cl and higher anion gap at 14 DIM. No differences were observed in body weight or body condition score due to DCAD or Ca. Prepartum dry matter intake (DMI) was lower for NEG compared with NEU and lower for 1.8% compared with 1.3% Ca. Postpartum DMI was not different among treatments. An interaction was observed for DCAD and DIM due to higher milk yield after 45 DIM for NEG compared with NEU. No differences were observed in milk component percentage or yield among treatments. There was an interaction of DIM and Ca for milk urea concentrations, which were higher at 5 wk and lower at 6 wk for 1.3% Ca compared with 1.8% Ca. These results suggest that feeding NEG prepartum alters plasma and urine mineral concentrations compared with feeding NEU and supports increased milk yield after 45 DIM. Feeding 1.8% Ca prepartum only improved plasma Ca at 1 DIM. Feeding either NEG or 1.8% Ca reduced DMI prepartum compared with NEU or 1.3% Ca.