Circulating antibodies following skin allograft rejection have been demonstrated by several investigators (1-8). There is little evidence, however, that these antibodies are important factors in the mediation of graft rejection (9). The dose-response and temporal relationships between allograft rejection and antibody formation have received relatively little attention (3, 10). The experiments reported here were designed to study these relationships with a sensitive isotopic antiglobulin technique ( 11). Some grafts were removed prior to rejection to determine the length of time a skin graft had to be present on an animal to stimulate antibody formation and to determine the effect of the presence of the graft on antibody level.Materials and Methods. Mice of three inbred strains were tested. B lO.D2/JaX (H2-d) mice were tested with C57B1/10 (H2-b) skin grafts, which differ only a t the strong H-2 histocompatibility locus. BALB/cAn (H2-d) mice were tested with C57B1/6N (H2-b) skin grafts, which differ at both H-2 and non-H-2 histocompatibility loci. Single, orthotopic skin grafts (1 cm2) were made on the back and were held in place with 9-mm Autoclips (Clay-Adams, New York) . After grafting, all animals were housed separately. They were divided into groups of 5-10 animals, and grafts were either left intact or were removed at days 2, 3, 5, 7. and 9. The clips were removed from the remaining grafts at day 7. The grafts were observed daily for visible and tactile signs of rejection.The mice were bled from the retro-orbital sinus before grafting and at 3-4-day intervals thereafter. Aliquots of serum were then stored at -20" until tested. The isotopic antiglobulin technique ( 11) was used to test the sera for antibody activity. All of the sera from an individual animal were tested at the same time. All sera were tested in duplicate or triplicate. The C57B1/6 ascitic leukosis (EL-4) was used for the target cells in all tests. Target cells (5 X lo5 to 1 X lo6 in 0.1 ml) were incubated with 0.1 ml of 1:s dilution of serum for 30 m h at room temperature in 10 x 75-mm siliconized test tubes. They were then washed seven times with 1 ml of Verona1 buffered saline, pH 7.4, containing 10% fetal bovine serum (VBS/lO% FBS). The cells were then incubated with 2 pl of 1251-antiglobulin for 15 min a t 0". The cells were again washed five times and then transferred to plastic counting tubes. Cells were counted for 1251 gamma emissions.A considerable range of antiglobulin u p take (1251 cpm) was noted when the results of the pregraft sera of the individual animals were compared. Testing of serial serum samples from the same unimmunized animal, however, yielded 1251 cpm which varied by only 5%. To determine the time of appearance of antibody activity, the results obtained with each animal's sera after grafting were compared with its own pregraft level.