Hypoestrogenism in postmenopausal smokers has been suggested as the mechanism for the observed decreased risk of endometrial cancer and increased risk of osteoporosis. We have prospectively studied a well-matched group of smokers and nonsmokers and have evaluated their estrogen levels and compared them with existing data from the literature. We conclude that increased adrenal activity resulting in increased androgens, mainly androstenedione, is seen in postmenopausal smokers but that estrogen levels are not decreased. We hypothesize that in nonusers, unlike in users of estrogen, smoking is not associated with changes in estrogen levels and that other mechanisms must be responsible for the epidemiologic observations seen.