Introduction: Metabolic syndrome (MetS) can lead to harmful effects on cardiovascular health. Assessing whether links between perceived neighborhood social environment (PNSE) and MetS are mediated by physical activity (PA) levels will inform interventions to reduce population-level mortality attributed to MetS. Hypothesis: PNSE will be both directly and indirectly related to MetS severity via PA. Methods: This study utilized baseline data from the Jackson Heart Study, a prospective, community-based cohort of African American adults in Jackson, MS (total sample=3082; mean age 53.3 years; 64% female). PNSE variables included neighborhood violence, problems (higher value=more violence or problems), and social cohesion (higher value=more cohesion). MetS severity Z-score was the outcome (higher score indicates worse MetS). Validated self-reported total PA, based on active living, sports, and home/yard indices, was tested as the mediator. Linear regressions with bootstrap-generated 95% bias-corrected confidence intervals ( BC CIs ) were estimated to test for significant unstandardized indirect effects, controlling for covariates (age, education, smoking status, alcohol consumption, and diet). All analyses were stratified by sex. Results: In women, neighborhood violence, problems, and social cohesion were indirectly related to MetS via PA (B for indirect effects=.04, 95% BC CI =.01, .08; .03, 95% BC CI =.01, .06; and -.03, 95% BC CI =-.07, -.01, respectively). That is, higher violence and problems were related to lower PA (B=-1.06, 95%CI=-1.76, -.36; B=-.79, 95%CI=-1.26, -.32, respectively), while higher social cohesion was associated with higher PA (B=.86, 95%CI=.15, 1.58). In turn, higher PA was associated with lower MetS. All PNSE factors were directly related to MetS (p<.05). In men, neighborhood violence, problems, and social cohesion were indirectly related to MetS via PA (B for indirect effects=.06, 95% BC CI =.02, .12; .04, 95% BC CI =.01, .08; and -.06, 95% BC CI =-.11, -.02, respectively). Again, higher violence and problems were related to lower PA (B=-1.65, 95%CI=-2.58, -.71; B=-1.02, 95%CI=-1.64, -.40, respectively), while higher social cohesion was associated with higher PA (B=1.58, 95%CI=.64, 2.52). In turn, higher PA was associated with lower MetS. No PNSEs were directly related to MetS. Conclusions: Consistent with what we expected, higher perceived neighborhood violence and problems were positively related to MetS, and vice versa for social cohesion. These associations may be explained in part by PA levels. The data showed that indirect effects by PA were more pronounced in men than women. Future interventions to reduce MetS attributed to neighborhood issues should consider PA promotion tailored to women and men in conjunction with community efforts to reduce neighborhood violence and problems and encourage cohesive neighborhoods.