The activation of N-nitrosobis(2-oxopropyl)amine (BOP) by pancreas acinar and duct tissue from Syrian hamsters and MRC-Wistar rats in vitro and in vivo was measured in terms of the production and repair of DNA damage. Hamsters were given BOP (1 X 10 mg/kg, s.c.). DNA single strand breaks (SSB) were measured over 2 weeks. Significantly more SSB were present in duct than in acinar tissue. Their persistence in the duct fragments was due to a slower rate of repair. In a related experiment, duct fragments were isolated from BOP-treated (1 X 10 mg/kg, s.c.) hamsters 24 h after exposure and cultured for 6 days, or were isolated 7 days after exposure. The extent of DNA damage was comparable in the two groups, indicating that the repair process(es) were still operative in cultured cells. Isolated duct fragments were exposed to either BOP or N-nitrosomethyl(2-oxopropyl)amine (MOP) in vitro. MOP produced significantly more DNA damage than BOP even at a 5-fold lower dose. This is consistent with the greater carcinogenicity of MOP in the pancreas. BOP produced significantly less DNA damage in the rat pancreas than in the hamster pancreas. The rate of repair was at least twice as fast in the rat pancreas as in the hamster pancreas. There did not appear to be any preference for acinar or duct tissue in rats as there was in hamsters. This procedure was validated in rats by the use of the rat pancreas carcinogen azaserine, which only produced DNA damage in rat acinar tissue.