Obesity risk reflects genetic susceptibility and social exposures. Perceived ethnic-based discrimination is a chronic stressor that may shape body mass index (BMI) through biobehavioral pathways. Polygenic risk scores (PRS) for BMI may correlate with or modify exposure to discrimination, and these associations may differ by sex and female reproductive factors. We estimate associations between a PRS for BMI (11% incremental R2) and perceived ethnic discrimination to evaluate gene-environment correlation (rGE) and gene-environment interaction (GxE), overall and by sex. Data were drawn from 8,593 unrelated adults in the HCHS/SOL baseline examination (2008–11).  Perceived ethnic discrimination was assessed via summative score of two questions about its frequency (to self & others) and standardized (mean=3.63, range 2-8). Linear regression models accounted for center, sociodemographics, and genetic ancestry; female-specific models additionally included reproductive factors. The sample’s mean age was 47 years (range 18-76); 57% female, and 42% living with obesity. Higher standardized PRS and discrimination scores were independently associated with higher BMI when jointly modeled (β=5.43, 95% CI=5.12-5.74; β=0.36, 95% CI=0.14-0.58). PRS for BMI was significantly associated with perceived discrimination (β=0.11, 95% CI=0.05-0.17), supporting rGE. A PRS x discrimination interaction was not statistically significantly associated with BMI. PRS remained significantly associated with BMI in both males (β=4.68, 95% CI=4.25-5.11) and females (β=5.93, 95% CI=5.5-6.36). PRS was not associated with discrimination among males, but was significant among females (β=0.67, 95% CI=0.38-0.96). In female models including reproductive factors, model fit improved and support for rGE remained. Findings support rGE over GxE, with sex-specific patterning, advancing understanding of how genetic and social processes jointly shape obesity risk in diverse populations.