We discovered by chance that the R28 T cell hybridoma has dual specificity. It responds to a peptide derived from ribonuclease presented by cells displaying Ak molecules and it reacts, in the absence of added antigen, to cells expressing Ak complexes with a single amino acid substitution at position 69 of the alpha chain. Modelling and functional studies suggest that residue 69 is a peptide contact residue, prompting the hypothesis that R28's alloreactivity is a cross-reactive response to an unknown peptide bound in the 'groove' of the mutant Ak complex. In this report, we employ a competition assay to confirm that this alloresponse involves a groove-binding peptide, demonstrate that this peptide derives from or depends on fetal calf serum and exploit a panel of antigen-presenting cell lines--each displaying an Ak complex with a different position 69 substitution--to establish that the alloresponse is not just a heteroclitic response to ribonuclease, itself. We speculate that much of the alloreactivity against murine class II molecules that is revealed in vitro may prove to be directed at bovine serum-derived peptides, suggesting that in this context, alloreactivity is a misnomer.