The precise role of breast milk leukocytes is unknown. We therefore studied the cellular composition of breast milk and the activation of breast milk macrophages in mothers with a cow milk-allergic infant and in those with a healthy infant. Further, we sought to determine the influence of the cellular composition of mother's milk on the infant's risk of developing cow milk allergy. Thirty-six asymptomatic mothers (26 with atopic constitution) whose babies had challenge-proven cow milk allergy and 24 asymptomatic mothers (17 with atopic constitution) with healthy infants were recruited. Geometric mean ages of the infants were 3.2 mo (95% confidence interval [CI], 2.3 to 4.4) and 2.4 mo (95% CI, 1.6 to 3.7), respectively. After separation of the fat layer, breast milk cells were incubated with fluorescein-labeled MAb to CD antigens (CD14, 45, 3, 4, 8, 19, 23, and HLA-DR) and analyzed by flow cytometry. Breast milk samples, collected with a breast pump, were processed immediately. Cytospin preparations of milk samples, made after separation of the fat layer, were stained with May-Grunwald-Giemsa and examined with a light microscope. HLA-DR expression on breast milk macrophages was significantly lower in the mothers whose infant was allergic to cow milk, 58.3% (95% CI, 44.9 to 75.6), than in the mothers of a healthy infant, 86.9% (95% CI, 78.7 to 96.1), p=0.012 (ANOVA). There was also a significant difference in the total number of breast milk leukocytes between the mothers with an allergic child, 0.17 x 10(6)/mL (95% CI, 0.12 to 0.25), and those with a healthy child, 0.08 x 10(6)/mL (95% CI, 0.05 to 0.14), p=0.019 (Mann-Whitney U test). These results suggest impaired function of breast milk macrophages in mothers whose infants had cow milk allergy. They may also reflect decreased antigen presentation to the inexperienced T cells in the gut or on other mucosal surfaces of the suckling infant, leading to subsequent development of food or respiratory allergies, e.g. asthma.