Studies of the induction of mammary tumors by 7,12-dimethylbenz(a)anthracene in a rat model show that human chorionic gonadotropin (hCG) administration reduces tumor incidence in a manner comparable to that of a completed pregnancy. On the basis of their studies, Russo and Russo (Cancer Epidemiol., Biomarkers & Prev., 3: 353-364, 1994) have proposed that hCG treatment of young nulliparous women would reduce their breast cancer risk in a manner similar to that of a term pregnancy. As part of a population-based, case-control study of breast cancer among women ages 40 years or younger, we asked women whether they had received hCG injection as part of a weight loss regimen or as a component of infertility treatment. Participants in this study were 744 women newly diagnosed with breast cancer between July 1983 and December 1988 and 744 controls individually matched on birthdate (within 36 months), race (white), parity (nulliparous/parous), and neighborhood of residence. Forty-five cases and 65 controls reported exposure to hCG (multivariate odds ratio = 0.77, 95% confidence interval = 0.50-1.19). Risk was reduced significantly among women whose maximum nonpregnant body mass index was less than 27.5 kg/m2 but no reduction in risk was observed among more obese women. Although the odds ratios were reduced substantially for both nulliparous and parous women with maximum nonpregnant body mass indices less than 27.5, only the result for nulliparous women was statistically significant. These results are consistent with the effects proposed by Russo and Russo based on their animal model. Although not definitive, these results suggest that hCG may be a means for reducing breast cancer risk.