Anonymous surveys have been widely used worldwide to describe adolescent substance use yet cannot elucidate causal drug abuse predictors. Studies in the U.S. have generally found that anonymous and confidential surveys yield comparable levels of self-reported substance use, yet the effect of survey format on youth self-report has not been evaluated in other countries. The present study compared data from the confidential International Longitudinal Survey of Adolescent Health with anonymously collected survey data on alcohol and marijuana use among school-based youth in Mexico, Puerto Rico, and the U.S. The findings suggest that confidential surveys yield valid self-reports of adolescent substance use.