Human sera contain cytotoxic naturally occurring (CyNa) antibodies which discriminate between lymph node cells from mice differing only at the H-2 complex. Sera from three healthy subjects (normal human sera, NH sera) and one serum from a patient with multiple sclerosis reacted with cells expressing Db, Kd, Kk, and KP molecules, respectively. However, the following observations suggested that the binding specificity of these CyNa antibodies is to antigens that are distinct from the classical H-2 antigens: (i) the NH sera did not contain cytotoxic anti-HLA antibodies, (ii) redistribution (capping) of H-2 antigens did not induce resistance to lysis for CyNa antibodies, and (iii) individual variation was demonstrated in the expression of the murine lymphocyte antigens detected by the human CyNa antibodies. The reason for this variation appeared to be different for individual NH serum. A maternal effect influenced the expression of the murine lymphocyte antigen detected by one NH serum (anti-H2b). The differences detected by another NH serum (anti-H-2P) appeared to be inherited, as shown by progeny testing. We hypothesize that the human CyNa antibodies may be directed against antigens controlled or modified by murine viruses (milk borne or endogenous), whose expression is under the influence of the H-2 complex, and that their production might have been stimulated by the products of human genes homologous to murine viruses.Human and animal sera contain cytotoxic "natural" antibodies (CyNa Abs) against lymphocytes from individuals of other species. Such antibodies are designated as nonspecific, speciesspecific, or heterophile. Studies on CyNa Abs are scarce and do not provide a conclusive explanation of their origin and specificity. It is not known (i) whether such antibodies are triggered by exogenous antigens or altered-self structures, (ii) what their target molecules are, (iii) what the reasons are for the species and individual variability in their occurrence and specificity, and (iv) whether they have biological relevance (1-4). Studies on CyNa Abs are hampered by their low titers and avidity, temperature sensitivity, and possible polyvalent nature. Only scant data are available about human CyNa Abs in sera of nontransfused male blood donors (normal human sera; NH sera). Such sera were used as controls in studies with hyperimmune anti-HLA antisera against panels of murine lymphocytes and were found to be either negative (5) or weakly positive (6) and did not distinguish normal murine lymphocytes carrying diferent H-2 haplotypes. One NH serum was found that distinguished, after absorptions, chicken erythrocytes with different major histocompatibility complex antigens (7).We tested a large number of NH sera and sera from diseased individuals on a panel of lymph node cells from congeneic mouse strains differing only at the major histocompatability complex.