Heterosexual HIV epidemics are driven by sexual contact between partners (sexual networks), yet little is known about how partnership patterns influence HIV transmission dynamics, and individual infection risk, in specific settings. This dissertation describes heterosexual partnership patterns and their potential contributions to the spread of HIV among networks of sexual contacts at elevated HIV risk in Ho Chi Minh City, Vietnam. The first paper employs within-cluster resampling methods to identify population patterns of heterosexual partnership selection, defined by selected HIV risk factors, within these networks. Nearly one-third (31%) of sexual contacts were HIV-positive. Young and unmarried men and women generally selected other young and unmarried sex partners living within their same districts; a pattern likely allowing HIV to persist within these networks. Men and women injecting drugs and exchanging sex typically chose non-injecting and non-exchanging sex partners; a situation likely to fuel ongoing HIV transmission within these networks. Most importantly, those involved in concurrent partnerships (more than one partnership at the same time) tended to partner with others in concurrent partnerships. This pattern will allow HIV to spread more rapidly through the networks than a pattern of serially monogamous partnerships. The second paper compares patterns of personal risk-taking and heterosexual partnering among female sex workers and non-sex workers within the networks. Notably, HIV prevalence was more than three times higher among non-sex workers than sex workers. Non-sex workers' own risk taking behavior was low; however, they had a high frequency of unprotected sex with HIV-infected primary partners, likely resulting in their exposure to infection. In comparison, sex workers had a higher frequency of personal risk-taking behavior and a lower frequency of exposure to risky partners, likely contributing to their exposure to infection. These findings illustrate that different patterns of individual risk-taking and sexual partnering may influence women's HIV risk. In the absence of condom use, these findings imply that women, even those whose own behavior is low risk, may be at increased risk for HIV if they are sexually connected to risky partners, even if they have just one, long-term, primary partner.