The mouse H-2Kd gene gives rise to several transcripts by alternative splicing. In addition to encoding the known Kd antigen, it could thus encode at least one minor hypothetical Kd-like molecule, with a distinct NH2 terminus. The existence of this "24" product can be inferred from a cDNA clone which was previously isolated. We have engineered both this cDNA and its canonical counterpart into a eukaryotic expression vector. After transfer of these constructs into mouse fibroblasts, we obtained cells expressing either one of the transcripts, but not both. In cytotoxicity tests, we found no expression of the "24" product on the cell surface, nor did we obtain any clue concerning its function. In contrast, cells which express Kd antigen, but none of the possible Kd-like molecules produced by alternative splicing, were functional in all aspects examined. We conclude that alternative splicing does not contribute to the known function of the Kd antigen.