The tobacco-specific N-nitrosamines (TSNA) have been implicated in oral cancer. However, except for one study using rats, no study has shown the ability of TSNA in inducing oral tumours in experimental animals. We have studied the carcinogenic potentials of N'-nitrosonornicotine (NNN) and 4-methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) in mice and hamsters, wherein the nitrosamines were administered on the tongues of the mice and the cheek pouches of the hamsters to simulate the exposure conditions of humans. It was observed that in Swiss and BALB/c male mice, both NNN and NNK induced tumours of lung, forestomach and liver. However, no oral tumours were induced in mice. The effect of vitamin A depletion was tested in Swiss male mice. It was found that a low vitamin A status did not alter the percentage incidence of tumours induced by both nitrosamines to a significant extent. In the studies using Syrian golden hamsters, long-term treatment of NNK to hamster cheek pouch induced tumours in the lung, liver, stomach and cheek pouch. Subsequently, the effect of hydrogen peroxide (H2O2) on NNK-induced carcinogenicity in hamsters was studied. It was observed that simultaneous administration of NNK and H2O2 to the animals increased the incidence of cheek pouch tumours. Another pertinent observation was that even when a small initiator dose of NNK was given followed by the application of H2O2, a very significant increase in the tumour incidence was observed. This observation suggests that H2O2 could act as a promoter to NNK-induced carcinogenesis. In conclusion it may be stated that both NNN and NNK do not show any strain or species specificity. They failed to produce tumours at the site of application in mice but in hamsters few cheek pouch tumours were seen or were induced when NNK was applied alone. The cheek pouch tumour incidence increased when H2O2 was given concurrently or when applied for a long period after a low initiator dose of NNK was administered in the cheek pouch.