HIV testing rates vary by race and ethnicity. Whether social capital indicators are related to HIV testing and whether these associations differ by race or ethnicity is unknown. Multivariable analysis was used to examine whether social capital (collective engagement and civic and social participation), including social cohesion (trust in neighbors, neighbors willing to help, feelings of belongingness) were associated with testing for HIV in the past 12 months. Participants were white, Black or African American, and Hispanic/Latino adults ages 18 to 44 (N = 2823) from the general population, in Philadelphia, PA who participated in the Southeastern Pennsylvania Household Health Surveys 2010 and 2012. Overall HIV testing in this sample was 42%, and was higher among women, and Black compared to white people. Mean social capital scores were significantly highest among whites. Greater trust in neighbors was associated with lower odds of testing for HIV (adjusted Odds Ratio[aOR]:0.61, 95% CI = 0.49–0.74), and this relationship varied by race/ethnicity, with stronger inverse associations among Hispanic/Latino (aOR = 0.43, p < 0.001) and white adults (aOR = 0.50, p < −0.001) than among Black adults (aOR = 0.75, p < 0.05). Greater neighborhood belongingness (aOR = 1.31, 95% CI = 1.11–1.54) and working together to improve the neighborhood (aOR = 1.33, 95%CI = 1.03–1.73) were associated with higher odds of testing for HIV. Different indicators of social capital were associated with higher as well as lower odds of testing for HIV. These patterns did not vary statistically by race or ethnicity. HIV testing prevention interventions will need to address social capital in design and implementation strategies.