A plenitude of research on juvenile recidivism exists within the criminological literature, and some scholars have suggested using the Latino paradox and racial invariance thesis to make sense of racial and other disparities in recidivism. However, there is an extremely limited body of literature that tests one or both of these theories on juvenile recidivism, and the research which does exist is limited in its generalizability. To address this gap, we use statewide data from Virginia’s Department of Juvenile Justice over 5 years to test the Latino paradox and racial invariance thesis. Furthermore, given the nested nature of the data (juveniles within counties/independent cities), we merged county-level data from the Virginia State Police and American Community Survey into the data and used two-level hierarchical generalized linear models to analyze the data. Our findings largely supported the Latino paradox and offered some support for the second interpretation of the racial invariance thesis. Implications for theory and recommendations for public policy are discussed.