The exposure of rats to the carcinogen 2-acetylaminofluorene (2-AAF) results in the accumulation of DNA-damaging adducts. The inability of cells to repair such damage adequately is a putative causal event in chemical carcinogenesis. It has been shown that one cellular response to DNA damage that leads to DNA repair is poly(ADP-ribosyl)ation of nuclear proteins. To examine the possible existence of an altered poly(ADP-ribosyl)ation response to 2-AAF-mediated damage of rat liver DNA, tissue ADP-ribose polymer levels were determined during various stages of 2-AAF-mediated carcinogenesis. 2-AAF was administered to rats in a discontinuous feeding regimen comprised of five consecutive cycles, each cycle consisting of 3 weeks on 2-AAF diet followed by 1 week of recovery on a control diet without 2-AAF. During cycle one of 2-AAF administration, rat liver ADP-ribose polymer levels increased 3-fold over that found in livers of rats fed only the control diet. In contrast, when rats were administered the non-genotoxic liver mitogen 4-AAF for one cycle, no significant elevation occurred in ADP-ribose polymer levels. Elevated ADP-ribose polymer production was also observed during cycles two and three of 2-AAF administration. However, during cycles four and five of 2-AAF administration, a period when rats administered 2-AAF acquire a high risk for hepatocarcinogenesis, an altered pattern of ADP-ribose polymer production occurred in rat livers. ADP-ribose polymer levels in these rat livers remained low, similar to levels found in control rat livers, despite the administration of 2-AAF. When the livers from rats fed either one or five cycles of 2-AAF were analyzed for possible decreases in the levels of tissue NAD+, the substrate for poly(ADP-ribose) polymerase, no changes in relative abundance were found. In addition, analysis of poly(ADP-ribose) polymerase activity showed no decrease at five cycles of 2-AAF administration. These results indicated that at late stages of 2-AAF-induced hepatocarcinogenesis, 2-AAF does not induce an expected increase in ADP-ribose polymer levels, and suggested that significant changes in DNA repair may occur at a time just preceding an increased risk for developing liver cancer.