The HLA region controls the major antigen system responsible for graft rejection and alloimmunization. More fundamental to the immune response, it controls the individual genetic elements regulating the interactions responsible for cellular immune functions. These HLA genetic determinants operate as permissive signals. When appropriate epitopes are available for presenting antigen to T cells, the effect is to facilitate response. If effector cells are activated, immune response is initiated. If suppressor cells are activated, immune response may be inhibited. Another genetic system separate from HLA presumably contains the genetic information that provides diversity for the T cell antigen receptor. Multiple nongenetic factors may also influence the immune response, among them prior alloimmunization via transfusion or pregnancy, viral infections and potentially a variety of other transfusion-related factors. Exposure to HLA and other tissue antigens has in general no clinical benefits and potential major adverse effects for an immunocompetent patient. Subsequent rejection of grafts of haematopoietic stem cells or of other tissues can result, as well as refractoriness to platelet and granulocyte transfusions. In the setting of profound immune suppression, infusion of viable lymphocytes can result in planned or inadvertent engraftment and a graft versus host reaction. The latter can occur even in the absence of measurable engraftment of haematopoietic cells. The apparent paradoxical effects of blood transfusion in aplastic anaemia, leading to marrow graft rejection, and in renal transplantation, leading to improved graft survival, reflect the lack of knowledge about which factors lead to Ir gene phenomena and which lead to Is gene phenomena. The observation that alloimmunization can result in specific immune suppression raises the exciting prospect of transplantation without the necessity of establishing a general state of immune compromise in the host. Since the effect of alloimmunization from transfusion or other sources is so unpredictable, it is difficult to ascertain whether transfusion is immunosuppressive in healthy individuals. Transfusion-related viral infections can lead to reversible immune compromise; much less is known about the effects of non-viral factors. Of these factors, HLA antigens in unaltered or in soluble form are probably important, but their relation to Ir and Is appears unpredictable. Evidence for transmission of AIDS via blood products has added to concern about transfusion-related immune suppression. Individuals at highest risk for transfusion-transmitted disease have received heavy exposure to multiple donor blood products. They are expected to exhibit the most pronounced non-specific immunological changes as well. If the aetiologic agent(s) responsible for AIDS can be identified, then the prevalence of seropositivity among those high risk patients can be ascertained and immunological effects unrelated to this disease can be identified. The problems in analysing this disease provide a more general caution about transfusion-transmitted infection. It is not clear that these infections need to be clinically overt to cause immunological alterations. An analysis of the immunological effects of HLA antigens or other non-infectious factors must therefore control as much as possible for potential infectious cofactors.