Infants born to atopic parents have been found to be at high risk of allergy development. The present study investigated the effect of a maternal milk-free diet during late pregnancy and lactation on the immune response and allergy incidence in at-risk and control infants. Atopic mothers were randomly allocated into an intervention group (n 12) or an unrestricted-diet group (n14) and compared with non-atopic mothers following an unrestricted diet (n 12). The intervention involved a maternal milk-free diet during late pregnancy and lactation. Infants were followed up for 18 months postnatally. A significant fall in maternal serum beta-lactoglobulin (beta-Lg)-immunoglobulin G (IgG) antibody levels (P < 0.05) was observed after a 7-week milk-exclusion diet. In maternal and cord serum samples the levels of beta-Lg-IgG and alpha-casein-IgG antibodies were significantly correlated (r 0.89, P < 0.0001 and r 0.71, P < 0.001 respectively). Higher levels of beta-Lg-IgG (P < 0.05) were observed in the cord serum samples compared with paired maternal serum samples. Single-blind allergy assessment by a paediatrician at 12 and 18 months showed that the infants born in the non-atopic group had a significantly lower allergy incidence compared with the infants born in the atopic group following an unrestricted diet (P < 0.008 and P < 0.02 respectively). The allergy incidence in the infants born in the atopic diet group was significantly lower compared with that of the atopic group following an unrestricted diet (P < 0.04). It was observed that the atopic nature of the parents significantly affected the allergy incidence in their children. A trend towards a beneficial effect of a maternal milk-free diet during late pregnancy and lactation was also observed in infants born to atopic parents.