Previous studies propose “transitivity” as a key network property. Our research suggests that when networks consist of relations that potentially carry major social costs, members of those networks may actively prevent transitivity (e.g., information flow), fostering in-transitivity. Using mixed research methods, we analyzed the social network structure of 175 female sex workers (FSWs) in post-socialist China. We identify four patterns of sex worker networks (i.e., dense, moderate, weak, separate relationships) depending on the degree of social cohesion between urban and rural networks. The distinct network configurations that we discovered result from a combination of factors: the level of hometown stigma (against female sex workers), age/work experience, and competitiveness in the sex industry. We conclude that triads trend toward in-transitivity when subgroups within social networks maintain competing cultural norms. As the first study mapping the social networks of female sex workers in China, this article demonstrates the women's aptitude for network management in order to prevent negative consequences of their engaging in stigmatized activities.