Prior research shows parental monitoring is associated with less substance use, but these studies have some limitations. Many examine older adolescents from White, Euro-American heritage, and cross-sectional studies are unable to test if parental monitoring decreases substance use over time. We address these limitations with longitudinal data of 2,034 primarily Latino preadolescents in Phoenix, Arizona, USA in 2004-2005. We use multilevel regression with multiple imputation of missing data. We find parental monitoring has beneficial, longitudinal effects on youth's substance use and related intentions, norms, and attitudes. Effects are invariant to gender or Latino ethnicity, except in the case of marijuana.