Caloric restriction has documented beneficial effects on numerous diseases including cancer, yet the mechanism(s) that accounts for these wide ranging benefits is unknown. Part of the difficulty in defining mechanisms has been the long-term nature of experimental protocols in which these beneficial effects have been observed and the inherent difficulty of investigating mechanisms in such studies. The experiments reported were designed: (1) to determine if caloric restriction would inhibit mammary carcinogenesis in a model for this disease process that is 35 days in duration; (2) to determine if progression from pre-malignant to malignant stages of mammary carcinogenesis was affected by caloric restriction; and (3) to explore whether the effects of caloric restriction were associated with changes in adrenal function. Mammary carcinogenesis was induced in female Sprague-Dawley rats by the i.p. administration of 1-methyl-1-nitrosourea (50 mg/kg body weight) at 21 days of age. Rats were randomized to one of four dietary treatment groups: ad libitum fed, or restriction of food intake to 90, 80 or 60% of the ad libitum intake. Rats were palpated for detection of mammary tumors and all mammary lesions excised at necropsy were histologically classified. Twenty-four-hour collections of urine were obtained at weekly intervals throughout the 35-day experiment. Urine was assayed for corticosterone by direct radioimmunoassay. Caloric restriction resulted in both a dose dependent prolongation of latency to palpable carcinomas (P < 0.01) and a reduction in final incidence of mammary cancer; the dose response was linear (P < 0.05). The percentage of pre-malignant mammary lesions in a group increased with increasing degree of caloric restriction, whereas the percentage of carcinomas decreased (P < 0.05). The level of cortical steroid increased linearly with increasing caloric restriction (P < 0.01) an effect that was not attenuated over time. Poisson regression analyses with the number of cancers per rat as the dependent variable, level of caloric restriction as the independent variable and urinary cortical steroid excretion as a co-variate were performed. These analyses indicated that the variation in cancers per rat, irrespective of the treatment group to which an animal was assigned, could be accounted for by urinary cortical steroid excretion (P<0.05); i.e. urinary cortical steroid excretion was an independent predictor of an animal's carcinogenic response. The data reported in this study support the use of a short term model to study the mechanism(s) by which caloric restriction inhibits mammary carcinogenesis and point to both a stage in the disease process, the conversion of pre-malignant to malignant cells, and a target tissue (adrenal gland) and chemical species (adrenal cortical steroid) that may be involved in mediating the protective effects of energy restriction. These data indicate the feasibility of identifying a chemical basis for the protective effect of caloric restriction that is independent of energy restriction per se and this, in turn, indicates that it may be possible to circumvent the practical problem of implementing a program of chronic energy restriction in human populations, yet still achieve the wide-ranging health benefits of such a program.