Introduction: Hypertension is a major risk factor for cardiovascular disease, stroke, kidney disease, and vascular dementia. Nearly two out of three Americans develop hypertension by age 70, and African Americans are disproportionately affected compared to non-Hispanic whites. The etiology of hypertension is complex, but there is an emerging consensus that both biological and social/psychosocial factors contribute to increases in blood pressure. Hypothesis: In this study, we assess whether the relationship between genetic factors and systolic and/or diastolic blood pressure is modified by social or psychosocial factors in up to 12,000 non-Hispanic whites and 3,000 African Americans from the Health and Retirement Study (mean age = 68). Methods: We conducted genomic region-level analysis using the Sequence Kernel Association Test (SKAT) to evaluate whether genetic variation at 34 known blood pressure loci (29 from published GWAS of European Ancestry, and 5 from trans-ethnic meta-analysis) were associated with systolic and/or diastolic blood pressure in each ethnic group, separately. Illumina HumanExome BeadChip data was used to evaluate the effects of rare exonic variants, while 1000 Genomes imputed data was used to evaluate common variants across the entire genomic region. For genomic regions that demonstrated a marginally significant effect on blood pressure (p<0.2), we used the Gene-Environment Set Association Test (GESAT) to evaluate whether social factors (education, parental education) or psychosocial factors (anger, stressful life events, social isolation, and depression) interacted with the genomic region to influence blood pressure. Results: Preliminary results indicate genomic region-level influences on blood pressure in non-Hispanic whites (ATP2B1, C10orf107, CACNB2, FES, MTHFR, ZNF652, and regions around rs1173711 and rs4373814) and African Americans (ATP2B1, FES) at p<0.05. Genomic region had a significant interaction with anger (C10orf107, PLCE1) in non-Hispanic whites, and both education (region around rs1458038) and anger (ULK4) in African Americans at p<0.001. Interaction analysis results for stressful life events, social isolation, and depression are forthcoming. Conclusions: New methods for integrating data on biological and social/psychosocial exposures are needed to more fully characterize the determinants of cardiovascular disease and elucidate potential mechanisms for cardiovascular disparities across ethnic groups.