The objective of this dissertation is to investigate how epidemiologic methods can be advanced through the use of social network analysis. Social network analysis (SNA) provides critical insights into the dynamics of health outcomes. Traditional epidemiologic methods have focused on the attributes and behaviors of individuals as the unit of analysis. However, a person’s risk for disease often depends on the risk factors and norms of the individuals within their social networks. This dissertation focuses on HIV infection to demonstrate the importance of social networks, as HIV is an inherently social disease that disproportionately affects particular socio-demographic groups in the United States (U.S.). , HIV incidence in the U.S. has remained generally stable in recent years, with the exception of the number of new infections among young Black Men who have Sex with Men (YBMSM). According to the CDC, HIV diagnoses increased by 87% among YBMSM from 2005-2014.1 Concurrently, only 3% of HIV funding is allocated for HIV prevention.2 Previous research has focused on individual-level risk factors as a means to explain the disparate trends in HIV prevalence and inform the design of HIV prevention interventions. This approach has been largely ineffective. Consequently, focus has now turned toward social network analysis (SNA). , SNA can be organized into two general analytic approaches: ego-centric (examination of the ties, attributes and local structure in one’s personal network) and socio-centric (examination of “whole” network structural characteristics). Social networks have been utilized to effectively promote HIV risk reduction and to recruit hard to reach populations and visualize where HIV transmission clusters occur. The potential of social network interventions, however, has not been actualized, primarily due to the dominance of individualism culturally and thus the domination of people with these views on review committees and funding agencies. The research described in this dissertation furthers HIV prevention research and epidemiologic methodology through the utilization and evaluation of two longitudinal social network based HIV interventions on the South Side of Chicago (SSC). Specifically, this dissertation employs the two SNA analytic approaches to:,Aim 1: Develop an HIV risk network metric to predict HIV acquisition over time. This metric (the “network viral load”) is based on the HIV viral loads of the sexual partners of an HIV negative individual.,• Hypothesis 1.1 Individuals with higher “network viral load” will be more likely to acquire HIV, and this relationship will be moderated by HIV risk behaviors, other network member attributes, and network position.,Aim 2: Describe social network stability among YBMSM ages 16-29 over an 18-month period and determine how factors related to social disorder, such as unstable housing, criminal justice involvement, and exposure to violence and resilience factors affect the composition of social networks over time ,• Hypothesis 2.1 Individuals with greater exposure to social disorder will experience higher rates of network instability, but this will be alleviated by resilience factors.,Aim 3: Determine whether a social network based HIV testing intervention in networks of individuals who were recently infected or recently diagnosed with HIV is a cost-effective strategy for locating people with undiagnosed HIV infection compared with the expanded HIV testing initiative in a hospital setting. ,• Hypothesis 3.1 The social network based testing will be more costly than expanded testing, but also more effective at yielding new HIV diagnoses.