Using data from the Chinese Health and Family Life survey, this study analyzed the prevalence of and risk factors for sexual harassment in China in the year 2000. It was the first study to use a general population sample to examine all types of harassment in an Asian country. The dataset was a stratified probability sample with 3,821 participants, and was nationally representative (apart from Hong Kong and Tibet) of China’s adult population aged 20–64. In total, 12.5% of all women and 15.1% of urban women reported some form of harassment in the past year. Among urban women age 20–45, most cross-sex harassment was not from supervisors or superiors (1.4%) but from coworkers and other peers (7.0%), strangers (4.6%), dates and boyfriends (3.6%), and others (2.6%). Multivariate analysis of risk factors for cross-sex harassment suggested that, despite its predominance in the Western literature on sexual harassment, the power differentials approach, focusing on male-female power differentials in patriarchal societies, was of modest utility. The results were more consistent with a more comprehensive routine activities approach borrowed from criminology, which emphasizes situational opportunity, perceived benefit to the harasser, and reduced costs for the harasser. The most striking result from the data represents the area receiving the least attention in the West, namely, the perpetrator’s perception of “benefit,” deriving from the victim’s inadvertent “signaling.”