Enzymatic cleavage of N-nitrosodimethylamine (NDMA) to nitrite (normally representing about 10% of the total metabolism in vitro) also produces methylamine in yields roughly equimolar to those of nitrite, suggesting that the 'denitrosation' pathway may be responsible for the previously unexplained detection of methylamine as a urinary metabolite of NDMA and, at least in part, for the recovery of less than stoichiometric amounts of dinitrogen in 15N-labelling experiments. We have now followed excretion of labelled methylamine by rats receiving 14C-NDMA as a possible index of the extent of in-vivo denitrosation. Correcting for the proportion of labelled methylamine recovered in the urine following its administration under the conditions used for NDMA, 2.5-10% of the NDMA metabolism in Fischer rats appeared to proceed by a methylamine-forming route. The results are consistent with the conclusion that the metabolism of NDMA is best viewed as a competition between two pathways, with denitrosation diverting a significant proportion of the clearance to a presumably deactivating metabolic route at the expense of the activating alkylation pathway responsible for carcinogenesis.