Contrary to expectation chickens did not readily elicit antibodies to IgA dimers when untreated human colostrum was used as antigen. When colostrum was fractionated by means of a column of 8% granulated agarose equilibrated with 10mM phosphate buffer, pH 7.2, a major and a minor fraction were obtained. The major or "1st fraction" consisted of two components with sedimentation coefficients of 10.9 S and 14.1 S, respectively. The minor or "2nd fraction" consisted of components of S values ranging from 2 to 6 and small amounts of 10.9 and 14.1 S material. When chickens were immunized with the "1st fraction" antibodies to dimeric IgA were produced. When the "1st and 2nd fractions" of the column were remixed and used for immunization of chickens, the immune response was as poor as when the chickens were injected with untreated colostrum. An immuno-depressing agent in colostrum was indicated. When rabbits were immunized with clarified human colostrum, antibodies against five antigens were elicited, one of the antigens being dimeric IgA. The immuno-depressing agent is therefore not universal. The purified agent suppressed antibody formation in chickens against the haemocyanin of Jasus lalandii. The "activity" is therefore not specific for IgA and the remaining four antigens in human colostrum. The purified component is a glyco-protein with a hexose content in excess of 10%. The derivatized sugars prepared from it were shown by gas liquid chromatography to be an equimolar mixture of galactose, mannose and fucose. The molecular weight (Mr) of the purified component was found to be 72,000 by sedimentation and diffusion and 80,000 by SDS page using Mr reference standards. The properties of the immuno-suppressor strongly suggest that it is the secretory piece of dimeric IgA.