Apparent digestibility and retention of Mn by dairy cows was used to compare 2 sources of Mn and to estimate Mn requirements. In experiment 1, Holstein cows at dry-off (60 d prepartum) were fed a basal diet with no supplemental Mn (43 mg of Mn/kg of dry matter) and received a daily bolus of 0 or 200 mg/d supplemental Mn from MnSO4 or from Mn-Met (6 cows per treatment) until parturition. Approximately 30 d before parturition, cows were moved to metabolism stalls for total collection of feces and urine. No differences were observed between Mn sources, but apparent absorption of Mn (6.4 vs. 2.3%) tended to be greater, and apparent retention of Mn (44 vs. 12 mg/d) was greater, for cows given supplemental Mn compared with control cows. In the second experiment, apparent Mn digestibility data from 8 experiments conducted with lactating dairy cows (39 dietary treatments and 160 observations) were combined with data from experiment 1. The regression equation of intake of digestible Mn on Mn intake (i.e., Lucas test) was as follows: intake of digestible Mn (mg/d) = -151 + 0.26 x Mn intake (mg/d). Based on that equation, Mn intake had to equal 580 mg/d to meet the metabolic fecal Mn requirement. The corresponding dietary concentration, assuming dry matter intakes of 21 and 12 kg/d for lactating and dry cows, respectively, were 28 and 49 mg/kg dry matter. These concentrations are approximately 1.6 and 2.7 times higher than those needed to meet the Mn requirements for lactating and dry cows, respectively, as calculated using the 2001 National Research Council dairy requirements model.