We studied the influence of thinness and smoking in 153 early postmenopausal women from a 3-yr period, comparing treatments of 1 or 2 mg estradiol with placebo. The baseline body mass index (BMI) was significantly associated with bone resorption (r = -0.26, P < 0.01 for urinary CrossLaps with 21% difference between extreme tertiles) with a consequent association between BMI and bone mineral density (BMD; r = 0.38, P < 0.001 for BMD of hip with 10% difference between extreme tertiles). A low BMI led to an increased rate of loss (r = 0.45, P < 0.01 for BMD of spine), whereas response to treatment with either 1 or 2 mg estradiol was independent of BMI. Smoking was associated with a 4% lower BMD at baseline compared with that in nonsmokers. This effect was additive with that of BMI. For smokers the increase in serum estradiol during hormone replacement therapy was only half of that in nonsmokers. For women treated with placebo or 2 mg estradiol, serum FSH levels were similar in smokers and nonsmokers, but during treatment with 1 mg estradiol, serum FSH was significantly less suppressed in smokers. This was mirrored in the BMD response, where smokers responded similarly to 2 mg estradiol and placebo as did nonsmokers, whereas smokers receiving 1 mg estradiol experienced only half the increase compared to nonsmokers. In conclusion, thinness and smoking are important risk factors for osteoporosis development, but are counteracted by hormone replacement therapy.