Some transplant centers consider paternal HLA antigens as unacceptable mismatches for mothers awaiting kidney transplantation. It is feared that a pregnancy may cause priming of the maternal immune response directed toward paternal HLA antigens. Should a woman receive an organ from a donor who shares those paternal HLA antigens, the risk of graft rejection might be increased. It is known that some women, as a consequence of pregnancy, develop antibodies specific for paternal HLA antigens. The purpose of the present study was to investigate whether a pregnancy can also prime the cellular immune response and whether this occurs in all cases. Frequencies of maternal cytotoxic T lymphocytes directed to paternal HLA antigens were evaluated in limiting dilution analysis assays and compared with those directed to third-party HLA antigens. Differentiation between naive and in vivo primed cytotoxic T lymphocytes was made by performing these assays in the absence and presence of anti-CD8, respectively. No difference in the frequency nor sensitivity to blocking by anti-CD8 was found when maternal cytotoxic T lymphocytes directed toward paternal HLA antigens were compared with those against third-party HLA antigens. However, more heterogeneous responses were detected against paternal HLA antigens. Therefore, paternal antigens that had been inherited by children were analyzed separately from the paternal antigens that had not been inherited. Furthermore, the presence of pregnancy-induced HLA antibodies was taken into consideration. Naive cytotoxic T lymphocyte responses were detected against paternal antigens that had never been inherited and those that had been inherited but had not induced antibody formation. In contrast, inherited paternal antigens that had induced HLA-specific antibodies in the mother gave rise to elevated cytotoxic T lymphocyte precursor frequencies, as compared with the response to third-party antigens. Also, the cytotoxic T lymphocytes were found to be more resistant to inhibition by anti-CD8, suggesting that these cells had been primed in vivo. These results suggest that not all paternal HLA antigens have to be considered as unacceptable mismatches. Only those individuals who share a paternal HLA antigen against which a mother has formed HLA-specific alloantibodies should be excluded from organ donation.