This study examines the separate relationships of public housing residence and subsidized housing residence to adolescent health risk behavior. Data include 2,530 adolescents aged 14 to 19 who were children of the National the Longitudinal Study of Youth. The author used stratified propensity methods to compare the behaviors of each group--subsidized housing residents and public housing residents--to a matched control group of teens receiving no housing assistance. The results reveal no significant relationship between public housing residence and violence, heavy alcohol/marijuana use, or other drug use. However, subsidized housing residents have significantly lower rates of violence and hard drug use, and marginally lower rates of heavy marijuana/alcohol use. The results indicate that the consistent, positive effect of vouchers in the current literature is not due to a lower standard among the typical comparison group: public housing. Future studies should focus on conceptualizing and analyzing the protective effect of vouchers beyond comparisons to public housing environments. (Contains 3 tables and 1 figure.)