In recent years, there has been an increasing interest in neighborhood environments as potential contributors to racial and socioeconomic health disparities. This trend reflects a growing recognition that individualistic explanations of health inequalities are both theoretically and empirically insufficient. Although neighborhood structural disadvantage has consistently been linked with increased rates of morbidity and mortality, the mechanisms through which neighborhood environments might get “under the skin” remain largely unknown. This dissertation contributes to the literature on neighborhoods and health by identifying potentially stressful and supportive dimensions of the neighborhood environment and testing their impact on both health outcomes and hypothesized physiological mediators. This dissertation begins by theorizing and constructing four nonsociodemographic measures of the neighborhood social and phyiscal environment: perceived stressors, observed stressors, social support, and participation. Since neighborhood effects have been documented for range of health outcomes, I use a widely accepted global indicator of health—self-rated health—to examine the relative effects of these different neighborhood dimensions on physical health. I find that perceived stressors have a strong negative effect on self-rated health, and appear to partially mediate the effects of neighborhood socioeconomic disadvantage. The second analysis examines the relationship between neighborhood stressors and support and cortisol, a commonly theorized physiological linking mechanism between stress and physical health outcomes. Using multilevel spline regression, I examine the effects of neighborhood characteristics on diurnal cortisol pattern. I find that individuals living in more stressful neighborhoods have lower overall levels of cortisol, characterized by blunted diurnal patterns. These results add to increasing evidence that long-term stress exposure can lead to hypocortisolism, which may have an important role in the pathophysiology of disease. In the final analysis, I examine the moderating role of gender in the relationship between obesity (measured by both BMI and waist size) and neighborhood socioeconomic, social and physical characteristics. Neighborhood disadvantage has a strong positive effect on BMI and waist size for women, but no effect for men. The results suggest that men and women respond differently to similar neighborhood environments in ways that are important for understanding the social causes of obesity.Ph.D.Social Work and SociologyUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/78869/1/rakarb_1.pdf