Allostatic Load (AL) is a cumulative index of physiological dysregulation. African Americans (AA) have traditionally had higher AL than their Caucasian American (CA) counterparts. Differences in psychosocial factors and negative health practices that disproportionately occur in certain groups may account for ethnic differences in AL. The main aim of this dissertation was to investigate if differences in psychosocial stressors, non-stress psychological variables and health practices mediated ethnic differences in AL (Aim 1). A secondary aim was to investigate methodological issues in the concept of AL. An AL composite score was calculated using a z-score method as well as the traditional scoring algorithm. Analyses from Aim 1 were conducted using each of these composite scores as an outcome to investigate if findings changed depending on the scoring method used (Aim 2). Using the sub-factors of AL as outcome variables, the tertiary aim was to investigate if psychological and health behavior mediators of ethnic disparities in the AL composite score also emerged as mediators of the individual sub-factor variables (Aim 3). Data were collected as part of a larger study investigating ethnic differences in cardiovascular risk factors. Participants included working AA (N = 75) and CA (N = 100), adult men and women. Participants completed questionnaires assessing demographics, psychosocial stressors, non-stress psychological variables, and health practices. Biological data were collected as part of an overnight hospital stay. The covariates age, gender and socioeconomic status (SES) were held constant in each analysis. Findings showed significant ethnic differences in AL, such that AA had higher AL than their CA counterparts (p < 0.01). Non-stress psychological variables and health practices were partial mediators of the relationship between ethnicity and AL (p's < 0.05). The experience of anger subscale from the Buss-Durkee Hostility Scale and the Pittsburgh Sleep Quality Index (PSQI) total score were significant individual mediators of the ethnicity-AL relationship (p < 0.05). Results were substantively the same regardless of the scoring algorithm used to create the AL composite score. Analyses using the sub-factors of AL were largely null. These results suggest that non-stress psychological factors and health practices play an important role in explaining ethnic differences in AL.