In this study, the anti-promoting effect of voluntary (wheel) exercise on 7, 12-dimethylbenz(a)anthracene (DMBA)-induced mammary tumorigenesis was investigated. All rats were fed high fat diets (23% of calories as fat) to mimice the typical western diet. Two doses of DMBA were used to determine if the antipromoting effects of exercise were dependent on the strength of the initiating agent. In addition, tumor estrogen receptors were assayed to determined whether exercise, through an estrogen-suppressing mechanism, selects for estrogen receptor-negative tumors. Female Sprague-Dawley rats were fed a semi-purified 23% fat (corn oil) diet (AIN-76A) and, on day 50 of age administered DMBA by gavage at 5 or 10 mg/rat. Rats were then randomized into 4 groups (n = 30) as follows: 1) low DMBA/sedentary; 2) low DMBA/exercise; 3) high DMBA/-sedentary; and 4) high DMBA/exercise. Active rats were placed in wheel-cage units, which allowed voluntary access to an activity wheel for 133 (low DMBA) and 77 (high DMBA) days, respectively, Sedentary rats were placed in conventional cages. Both active groups exhibited significantly lower total tumor numbers than their sedentary controls: 75 vs 102 (low DMBA) (p < 0.05) and 90 vs 160 (high DMBA) (p < 0.001). Compared to sedentary controls, latency was significantly lengthened in the low but not the high DMBA active groups; multiplicity, in contrast, was significantly decreased in the high, but not the low DMBA exercised group. Exercise had no effect on overall tumor incidence. When segregated into exercise tertiles, total tumor active compared to the least active tertile, particularly in the high DMBA group.(ABSTRACT TRUNCATED AT 250 WORDS)