T-cell-deficient mice, either anti-thymocyte serum treated or nude mice, were immunized with hapten (4-hydroxy-3,5-dinitrophenyl acetic acid, NNP) conjugates of syngeneic, allogeneic, or xenogeneic erythrocytes. Immunization with syngeneic conjugates led to a stronger anti-NNP response than immunization with allogeneic or xenogeneic conjugates. A study of congenic mouse strains suggested that a prerequisite for this effect was that immunogenic erythrocytes and responding animals shared H-2-controlled characteristics. F1 hybrid erythrocyte conjugates injected into F1 hybrid mice behaved like other syngeneic erythrocytes. The same erythrocyte conjugates injected into either parental strain induced a weak response indistinguishable from the response to allogeneic erythrocyte conjugates. Parental erythrocyte conjugates injected into F1 mice induced an anti-NNP response that was significantly lower than the response to F1 erythrocyte conjugates but significantly higher than the response to allogeneic conjugates. The response of normal mice to syngeneic erythrocytes was weaker than the response of T-cell-deficient mice, which could have been caused by suppressor T cells. Their response to allogeneic conjugates was higher than the response of T-cell-deficient mice and the response to xenogeneic conjugates higher still. This was probably due to allo- or xenoreactive helper cells.