The effects of dietary vitamin A on forestomach tumorigenesis during the total stage of the initiation and postinitiation periods and during the postinitiation stage were evaluated in ICR/Jcl mice treated with either high or low doses of benzo(a)pyrene (B(a)P). In experiment 1, the animals were initiated with a high carcinogenic dose of B(a)P to a total of 20 mg, while in experiment 2 the animals were treated with a low dose of B(a)P to a total of 2 mg. A control group of animals received no carcinogens. Five different dietary levels of vitamin A supplements were used in each experiment and in the control study. In experiment 1, a high incidence of tumorigenesis was observed in every group, with 74% to 96% developing papilloma and 19% to 46% developing carcinoma. In experiment 2, the incidence of tumorigenesis in the high-dose vitamin A groups, including those given during the total and postinitiation stages, was found to be significantly reduced at 7.4%, compared with that in the low-dose vitamin A group of 57.7% (P < 0.05). These results suggest that a high dietary level of vitamin A can reduce the incidence of tumorigenesis when low carcinogenic dose levels of B(a)P are given in both the total and postinitiation stages.