Putative risk factors (DNA damage) and risk modifying factors (DNA repair and cell proliferation) were examined in an experimental mouse model in which treatment with dimethylhydrazine (6.8 mg/kg DMH i.p. once weekly) for up to 20 weeks induces colon tumours in a site specific manner with 0, 43 and 87% of animals having proximal, mid and distal colon tumours respectively at the highest cumulative dose. Levels of the pro-carcinogenic DNA adduct, O(6)-methylguanine (O(6)-MeG), in colonic DNA were found to vary with time after final treatment and with location within the colon but not with total DMH dose. O(6)-MeG levels were generally lowest in proximal colon DNA and highest in distal colon DNA. Steady state O(6)-MeG levels were obtained at the highest cumulative DMH dose with O(6)-MeG levels in mid and distal colon DNA being 5 and 10 times higher those in proximal colon DNA. O(6)-alkylguanine-DNA alkyltransferase (MGMT) activity, and cell proliferation indices in the colon were also found to vary with time after final treatment but not with either location within the colon or total DMH dose. O(6)-MeG levels, MGMT activity and cell proliferation indices at specific time points as well as basal MGMT activity were not associated with differences in tumour yield within the colon. However tumour yield was associated with the cumulative amount of O(6)-MeG present in DNA over the treatment period and with the treatment induced cumulative increase in cell proliferation, particularly within regions of the colon crypt where stem cells reside but not with cumulative changes in MGMT activity. Results are consistent with an increased cancer risk arising from an increased mutation load in the target stem cell population due to increased adduct formation/persistence and cell proliferation but also suggest that other cell specific factors may help to determine tumourigenic response.