The transmethylation catalysed by HMT (EC 2.1.1.8)* has been considered as absolutely specific for histamine as acceptor substrate. In this investigation Nα-MH, Nα,NαDMH, spinaceamine and synthetically prepared 4-[(2-amino-ethylmercapto)-methyl]-imidazole could be identified as further methyl-group accepting substrates (optimum substrate concentration ~ 1 mM), but the yield of extractable 14C-labelled methylation products was never greater than 21 per cent of that of histamine. The 3 per cent methylation of Nα,Nα-DMH was considerably smaller than that of 33 per cent reported in the literature. This discrepancy was resolved and found to be ascribable to an inappropriate product extraction procedure used in the former experiments. When Nα-MH and Nα,Nα-DMH were the substrates, the corresponding products were isolated by t.1.c. in four different solvent systems and identified to be Nτ,Nα-DMH and Nτ,Nα,Nα-TMH. Thus HMT catalysed in all cases a uniform methyl of the Nα-nitrogen atom of the imidazole nucleus. The investigation of a series of various substituted imidazole compounds revealed that a methylation of the ring system had to be considered, if it was not substituted in the Nτ-, 2- or Nπ-position and if it carried a 4-substituent with a strong basic aminogroup, whereas substitution in the ring 5-position seemed to be of minor importance. Furthermore H1-receptor antagonists, H2-receptor antagonists, the non-imidazole H2-receptor agonist dimaprit, as well as the enzyme inhibitors aminoguanidine, tranylcypromine, pargyline and nicolinamide, were not methylated under the catalysis of HMT. The evidence for a less high substrate specificity of HMT may influence the relevance of histamine determinations using this enzyme: caution seems necessary.