Simultaneous colostrum (C) and plasma samples (P) from 14 women, 1 to 5 days postpartum, were examined. Total IgD and specific IgD antibodies to β- lactoglobulin, bovine serum albumin, Bermuda grass, and a-gliadin were measured by solid phase radioimmunoassay. The geometric mean concentrations of IgD were 35.8 (range 2.2-410) µg/dl for colostrum and 591.3 (range 72- 4100) µg/dl for plasma. Six subjects had a specific IgD antibody C/P ratio more than 10-fold greater than the total IgD C/P ratio, suggesting enhancement of antibody to a specific antigen in the mammary gland. All six had C/P ratios suggestive of local enhancement of IgD antibody to Bermuda grass, and two met this criterion for enhancement of IgD antibodies to β-lactoglobulin, bovine serum albumin, or a-gliadin. Specimens for these studies were obtained during the peak grass pollen season. Seventeen additional subjects were studied to compare total IgD in colostrum and plasma with total IgG and serum albumin. The mean C/P ratio for IgD (0.055 ± 0.015) exceeded the C/P ratio for total IgG (0.015 ± 0.003) or total albumin (0.020 ± 0.002). For 14 of 17 subjects the colostrum/plasma ratio for IgD exceeded the C/P ratio for albumin or IgG. Data were transformed logarithmically and correlation coefficients calculated. For albumin versus IgG in colostrum, there was a strong correlation, r = 0.865, p = 0.001. This was different from albumin versus IgD, r = 0.489, p = 0.046 and from IgD versus IgG, r = 0.556, p = 0.020. These analyses support a different mechanism of entry of IgD into milk compared to IgG or albumin. These studies also suggest that IgD antibodies may participate in responses of the mucosal immune system.