Maternal vitamin B12 deficiency results in low breast milk B12 and poor infant status. Both maternal and infant status could improve by maternal supplementation but the effective dose is not known. A dose‐response pilot study was designed to answer this question. Lactating Guatemalan mothers were screened to identify those with deficient (<148 pmol/L) and marginal (149‐225 pmol/L) serum B12 (69% of those tested). From these 18 mother‐infant pairs were stratified at 2 ± 0.5 mo postpartum by baseline serum B12 and randomized to 3, 100, 250, 500, 750 or 1000 ug B12/d for 2 mo, and 16 pairs completed the study. Outcomes at 1 and 2 mo were maternal and infant serum B12, infant urinary methylmalonic acid (UMMA) measured by HPLC/MS/MS, and milk B12 using chemiluminescence after removal of unbound haptocorrin by cobinamide‐coated EAH sepharose gel. At baseline 62% of women consumed <EAR for B12, maternal serum B12 was inversely related to infant UMMA (r=‐0.442, P<0.06) , and milk B12 was related to maternal UMMA (r=‐0.638, P<0.02). After 1 mo there was a dose response in maternal serum B12 (r=0.77, P<0.001) and milk B12 (r=0.78, P<0.001) but not infant serum B12 or infant urinary MMA. Results were similar after 2 mo. Only 1% of the 3 ug dose and 0.01‐0.05% of doses =100 ug/d would be transferred to exclusively breastfed infants. Supplementation of lactating women cannot increase breast milk B12 sufficiently to normalize infant status.