In two experiments, the effects of caloric restriction during the postinitiation phase of pancreatic carcinogenesis were evaluated. Male Lewis rats were given injections of azaserine at 14 days of age and weaned to the postinitiation test protocols at 21 days of age. In the first experiment, the caloric content of the diets was restricted by 10, 15, 20, and 30% of the intakes of the ad libitum-fed rats. A sixth group was fed diet ad libitum for only 5-6 h/day; i.e., they were "meal-fed". The development of putative preneoplastic lesions (henceforth termed foci) was evaluated by quantitative stereological (morphometric) analysis of the pancreas. Caloric restriction during the 4-month postinitiation phase resulted in a significant reduction in focal development beginning at 10% caloric restriction and increasing with more severe restriction. The caloric intake of the meal-fed group closely matched the caloric intake of the 10 or 15% caloric restriction groups and the focal response of the meal-fed rats was similar to the groups restricted in calories by 15 to 20%. In the second experiment, rats were initiated with azaserine and weaned to one of four groups: ad libitum; meal-fed; meal-fed for 2 months and ad libitum thereafter; or ad libitum for 2 months and meal-fed thereafter. Foci were evaluated at 2 and 4 months; neoplasm incidence and multiplicity were determined at 14 months postinitiation. Compared to the ad libitum group, the meal-fed group had significantly fewer foci at all times of evaluation and significantly fewer neoplasms. When rats were meal fed for 2 months and then switched to ad libitum feeding for the remainder of the experiment, the focal outcome at 4 months was similar to the group meal fed for all 4 months; and at 14 months the neoplastic outcome was intermediate between the ad libitum and the meal-fed group. Intervention in the ad libitum feeding regimen at 2 months by meal feeding for the remainder of the experiment resulted in a significant decrease in the focal and neoplastic development, as compared to the group fed ad libitum continuously. These two intervention groups were intermediate in response between the meal-fed and ad libitum-fed groups. These results indicate that the postinitiation phase of pancreatic carcinogenesis can be modulated by relatively simple dietary interventions such as moderate caloric restriction.