Disparity in HIV prevalence by race/ethnicity has been noted among men who have sex with men for almost 20 years. Research suggests that rather than individual risk behaviors, sexual networks play an important role in HIV prevalence disparity. This analysis uses data from the National HIV Behavioral Surveillance System collected at three time points using time-location sampling in 2004, 2008, 2011. We use Newman's coefficient to assess racial mixing among 1,207 observations. We found significant differences in HIV status across race groups in 2004 and 2008; yet in 2011, there were no significant differences of HIV status by race. Racial mixing across all races increased from 2004 to 2011; in other words, individuals were increasingly more likely to sexually partner outside their own race/ethnicities. Increased racial mixing may explain this convergence, although full social network studies are necessary in order to fully explain these findings.