Variation in the risk of breast cancer observed among women and among populations may be explained by variation in lifetime exposure to estrogens. The suppressive effect of exercise on estradiol levels in women is well documented, but it is unknown whether habitual (i.e. typical daily) physical activity has a similar effect. Epidemiological data suggest that physical activity is one of the few modifiable factors capable of reducing the risk of breast cancer in women. We investigated whether variation in the amount of habitual activity corresponds to variation in estradiol levels in women of reproductive age. One hundred and thirty-nine regularly menstruating women 24-37 years of age collected daily saliva samples for one complete menstrual cycle and kept a daily log of physical activity. Saliva samples were analyzed for concentration of estradiol. We observed a negative relationship between habitual physical activity and salivary levels of estradiol. Mean estradiol was 21.1 pmol/l in the low, 17.9 pmol/l in the moderate and 16.6 pmol/l in the high activity group (all pairwise differences were statistically significant at P<0.009). A strong association exists between physical activity and levels of estradiol among women of reproductive age. A modern lifestyle, characterized by reduced physical activity, may therefore contribute to a rise in the levels of estradiol produced during menstrual cycles and thus to higher cumulative lifetime exposure to estradiol, resulting in a higher risk of breast cancer.