An antibody was raised against a synthetic peptide (Ser-Glu-Asn-Tyr-Lys-Asp-Asn) corresponding to residues 290-296 of the cytochrome P450 enzyme, CYP1A2, of both rat and mouse. A cysteine residue attached to the N-terminus of the peptide during synthesis allowed coupling in a specific orientation via the thiol group to the carrier protein, keyhole limpet haemocyanin. Antiserum raised in rabbits bound specifically to CYP1A2 in the rat and mouse. To determine those amino acid residues involved in binding of the antibody, related peptides of various lengths were synthesised and the binding of the antibody was determined by an enzyme-linked immunosorbent assay. These studies show that the minimum epitope is the C-terminal tripeptide sequence, Lys-Asp-Asn. Other than in rat and mouse CYP1A2, this tripeptide is found as an internal sequence in a large number of proteins including bovine fibronectin, chicken gizzard myosin heavy chain, and the P450 enzymes, rabbit CYP3A6 and human CYP3A4, but the antibody did not bind to any of these proteins. However, the antibody did bind to yeast glucose-6-phosphate dehydrogenase in which the tripeptide sequence is the C-terminus. Antibodies raised against a truncated peptide (Tyr-Lys-Asp-Asn), representing the C-terminal half of the peptide, also bound to glucose-6-phosphate dehydrogenase, but failed to bind to CYP1A2; thus although the C-terminal region of the peptide 290-296 is strongly immunogenic, it appears that it is not this population of antibodies that binds to CYP1A2. As antibodies were found to bind strongly to the C-terminus of glucose-6-phosphate dehydrogenase, the C-termini of proteins as targets for anti-peptide antibodies were investigated further by immunising rabbits with four 5-residue peptides which represent the C-termini of the P450 enzymes, CYP1A1, CYP1A2, CYP2E1 and CYP2A6. The peptides were coupled to keyhole limpet haemocyanin through their N-termini via cysteine residues added to the sequences. All four antisera bound specifically to their respective target proteins, as demonstrated by immunoblotting using hepatic microsomal fractions from rat, rabbit and human. It is suggested that this method of antibody production could be of general use for the reliable production of antisera against proteins where their sequence at the C-terminus is known, and such antibodies can be highly specific as they do not bind to internal sequences.