I used multivariate ordinary least-squares regression to estimate premature mortality, adjusting for gender, marital status, education, mental illness diagnosis, race, and cause of death. After the analysis controlled for the other factors, race (b=2.5, p≤.001) and a mental illness diagnosis (b=4.5, p≤.001) were significantly related to premature mortality. The regression coefficients were used to calculate the predicted values of years of potential life lost (YPLL) for CSS and non-CSS decedents by race. The difference in YPLL between African Americans and whites was smaller among CSS decedents than among non-CSS decedents (2.1 years and 2.5 years, respectively). CSS decedents who were African American lost 15.8 years of life, compared with 13.7 years for white CSS decedents. Among non-CSS decedents, the YPLL for African Americans was 11.7 years, compared with 9.2 years for whites, for whom YPLL was lowest. [A table summarizing these results is available as an online supplement to this letter at ps.psychiatryonline.org.] As a final step, an interaction term was entered in the regression model to assess whether a mental illness diagnosis had a differential effect on YPLL by race. This interaction was not significant.