The objective of the study was to establish whether cholic acid (CHA) enhanced colonic tumor incidence in the early phase of carcinogenesis. Male, Sprague-Dawley rats (n = 180) were injected twice with azoxymethane (AOM) (15 mg x kg(-1) body weight x week(-1), s.c., given 1 week apart). Following the first AOM injection, animals were randomly assigned to two groups, control AIN-93G diet (CON) or control diet containing 0.2% CHA by weight (CHA). Three weeks after the first injection, 20 animals (10 animals/group) were killed and aberrant crypt foci (ACF) were enumerated. The remaining animals were further subdivided and animals randomly assigned to CON or CHA diets, creating four treatments: CON-CON, CON-CHA, CHA-CHA, and CHA-CON. After 3, 12, and 20 weeks (following the first carcinogen injection), the animals were killed and the number and crypt multiplicity of ACF enumerated. Macroscopic tumors were evaluated at week 20. Total ACF were not different between groups. Average crypt multiplicity and medium (4-6 crypts/focus) and large (> or = 7 crypts/focus) ACF were greater in CHA-CHA and CHA-CON compared with CON-CON and CON-CHA (p < 0.01). Transient exposure to CHA (CHA-CON) was sufficient to induce development of ACF with an accelerated growth phenotype and elicit a tumor-enhancing effect. CHA-CHA had the highest tumor incidence (82.8%, p < 0.05) followed by CHA-CON (56.7%, p < 0.05), and tumor multiplicity and number of tumors per rat in CHA-CON were similar to CHA-CHA (2.29 and 1.3 versus 2.33 and 1.9, respectively). Delayed intervention with CHA (CON-CHA) produced a tumor outcome similar to CON-CON (31 and 30%, respectively), it did not enhance colonic tumor incidence. Taken collectively these results suggest CHA was effective in enhancing colon carcinogenesis during early phases and ineffective in post-initiation phases.