Black and Latino populations have been disproportionately burdened by COVID-19 morbidity and mortality. Subsidized housing, crowding, and neighborhood poverty might be associated with increased COVID-19 transmission and play a role in observed racial and ethnic disparities, yet research is limited. Our study investigated whether these housing variables mediate the relationship between race and ethnicity and SARS-CoV-2 antibody seropositivity among New York City (NYC) adults. We analyzed data from a SARS-CoV-2 serosurvey (n = 1074), nested within the 2020 cross-sectional NYC Community Health Survey (June–October 2020). We defined SARS-CoV-2 seropositivity as either a positive blood test for SARS-CoV-2 antibodies or a self-reported positive test result. We used causal mediation analyses to test whether subsidized housing, crowding, and neighborhood poverty mediate a relationship between race and ethnicity and seropositivity. After controlling for potential confounding, we found elevated prevalence ratios of SARS-CoV-2 seropositivity among Black (APR = 1.74, 95% CI = 1.10–2.73) and Latino (APR = 1.58, 95% CI = 1.05–2.37) residents compared with White residents and for those living in crowded housing (APR = 1.48, 95% CI = 1.03–2.12) and high-poverty neighborhoods (APR = 1.54, 95% CI = 1.12–2.11) but not for subsidized housing. We observed statistically significant natural direct effects for all three mediators. While living in crowded housing and high-poverty neighborhoods contributed to racial and ethnic disparities in seropositivity the estimated contribution from living in subsidized housing was −9% (Black) and − 14% (Latino). Our findings revealed racial and ethnic disparities in seropositivity of SARS-CoV-2 antibodies among NYC adults. Unlike crowding and neighborhood poverty, living in subsidized housing did not explain racial and ethnic disparities in COVID-19.