SUMMARY Thirteen lots of horse antihuman lymphocyte globulin (ALG), produced using several different antigen sources and methods of preparation, were tested in the primate skin allograft assay system. Each ALG tested showed significant prolongation of skin allograft survival (range, 15-44+ days; average, 22.79 ± 8.93 days) compared to the control (9.95 ± 1.46 days). No antigen source, method of preparation, or route of administration produced consistently longer survival and none prevented the formation of specific antibody against the skin graft donors. I.v. administered ALG did not lead to the fixation of antiglomerular basement membrane (GBM) antibody as assessed by conventional immunofluorescent techniques and resulted in a lower incidence of antihorse IgG antibody production than that following administration by the s.c. route. One lot of ALG, however, caused fatal disseminated intravascular coagulation in four monkeys, only when administered i.v. A positive correlation between primate skin allograft survival and rosette inhibition (RI) titers was not demonstrated.