This work was designed to study the effect of the vitamin D content of human milk on the vitamin D status of exclusively breast-fed infants, and the relation between milk and maternal serum concentrations of vitamin D during the first month of lactation. Serum levels of calcium (Ca), phosphorus (P), magnesium (Mg) and 25-hydroxyvitamin D (25-OH-D) were determined in a racially heterogenous population of nursing women, between days 3 and 5 (L3), 15 and 18 (L15) and 30 and 45 (L30) post partum. The same parameters were determined in the serum of 1-month-old breast-fed infants. Maternal milk samples were obtained at L3, L15 and L30 and analysed for Ca, P, Mg, vitamin D and 25-OH-D content. Milk levels of Ca, P and Mg were found to be within the range previously described by other authors. No correlation was found between serum and milk levels of vitamin D and 25-OH-D in nursing mothers. The 25-OH-D concentration in milk was related to its vitamin D content and strongly correlated (P < 0·001) with the 25-OH-D levels in the serum of exclusively breast-fed infants. No significant changes were observed in maternal serum levels of 1,25-dihydroxyvitamin D (1,25-(OH)2D3) measured at L3 and L30, or between maternal and infant levels of 1,25-(OH)2D3 at L30. This study emphasizes the importance of the 25-OH-D content of maternal milk, in being primarily responsible for the vitamin D concentrations found in the serum of exclusively breast-fed infants. In contrast, serum levels of 1,25-(OH)2D3 measured in the breast-fed baby seemed mainly related to its calcaemia.