Data from large observational cohort studies of people with HIV have yielded equivocal findings on the relationship between substance abuse and HIV progression. A key problem is the inconsistent and varied definition of substance abuse employed in the analyses. To address such limitations, this dissertation approached substance abuse exposure measurement using latent variable techniques to better capture patterns and trajectories of substance abuse and to assess their effects on several clinical and public health HIV outcomes. Data derived from 6440 HIV infected and uninfected adults participating in a multi-site prospective cohort study, the Veteran's Aging Cohort Study (VACS), who were followed over four years. Three independent, conceptually related manuscripts were developed for this dissertation. The first employed a cross-sectional approach and discovered five empirical patterns of drug abuse (non-medically prescribed psychoactive drugs), the composition and prevalences of which were heavily influenced by HIV status. In the second manuscript, associations between drug abuse patterns and all-cause mortality were assessed. While results indicated no direct effects of drug abuse patterns on mortality risk among HIV infected and uninfected, the patterns interacted with other risk factors in important ways, suggesting that drug use influences mortality indirectly in this population. The third manuscript modeled drug and alcohol use patterns over four years to uncover trajectory classes and examined their role in determining HIV disease progression and medication adherence longitudinally. Eight trajectory classes within each disease status were detected, four of which replicated the cross-sectional patterns. Substance abuse remained stable among the HIV infected, but HIV uninfected respondents reduced use over time. Among the HIV infected, trajectory classes reflecting recreational and continued use of drugs and alcohol were associated with disease progression and lower medication adherence, while current marijuana use was associated with lower risk of severe immune dysfunction. In summary, results depicted substantial substance abuse in a midlife adult population suggesting needs for increased screening for substance abuse problems and expansion of age-appropriate treatment. At greatest risk for poor HIV outcomes were those continuing drug and alcohol use. Findings suggest more targeted medication adherence and treatment interventions for this group are warranted.