Because smoking is associated with an increased risk of osteoporosis, yet a decreased risk of endometrial carcinoma, a state of relative hypoestrogenism induced by smoking has been suggested. However, because previous data are unclear and do not reflect current trends in smoking intensity and estrogen prescriptions, we examined the estrogen profiles of postmenopausal women, by smoking status, both before and after oral micronized estradiol. Baseline levels of estrone, estradiol, estrone sulfate, and estrone glucuronide were similar in nonsmokers and smokers, but unbound (non-sex-hormone-binding-globulin--bound) estradiol was significantly lower in smoking women (p less than 0.05) and sex-hormone-binding-globulin--binding capacity was higher (p less than 0.001). After 1 or 2 mg of micronized estradiol, estrone and estradiol serum profiles were similar but unbound estradiol was significantly lower in women who were smokers (p less than 0.05). Serum estrone glucuronide rose with treatment but was indistinguishable in nonsmokers and smokers. However, maximum changes in serum estrone sulfate were greater in smokers after administration of estrogen, suggesting a hepatic effect. Urinary estrone glucuronide levels increased after 8 hours of oral estrogen but were similar in nonsmokers and smokers with the two doses. It appears that even moderate smoking, as studied here, induces significant changes in hepatic estrogen metabolism and is best reflected by alterations in serum estrone sulfate and sex-hormone-binding-globulin--binding capacity that result in decreased serum unbound estradiol. However, these changes do not appear to require increasing the estrogen dosage to achieve physiologic levels of estrogen in postmenopausal smokers.