A case-control study involving 331 patients with complete hydatidiform mole and 662 community controls matched to the cases on age and timing of pregnancy was conducted in Beijing, China. A history of a term birth was associated with reduced risk (odds ratio = 0.6, 95% confidence interval 0.4 to 0.9), with some evidence of further decrease with multiple births. Previous spontaneous abortions were not related to risk, although those with a prior induced abortion were at elevated risk, particularly if two or more abortions were involved (odds ratio = 2.8, 95% confidence interval 1.4 to 5.7). A history of having sought medical advice for infertility was associated with reduced risk (odds ratio = 0.5, 95% confidence interval 0.2 to 0.8), but those who reported use of herbal medicines during a first trimester of a previous pregnancy were at excess risk (odds ratio = 2.2, 95% confidence interval 1.3 to 3.6). In addition, a statistically significant trend in risk was observed with years of oral contraceptive use (odds ratio = 2.6, 95% confidence interval 0.9 to 6.9 for greater than or equal to 4 years of use). Dietary habits and family histories of cancer or trophoblastic disease were not related to risk in this study.