This analysis examines the impact of established predictors on probation failure utilizing a large randomly selected sample of adult probationers. Initial findings suggest that race, gender, location, offense severity as well as risk assessment scores significantly predict probation failure. This study then examines interaction effects between race and gender as well as race and offense severity. Results indicate such interactions may matter in studying probation failure, despite reason to be cautious about their interpretation. Importantly, the results of the interaction model suggest that the interaction between being an African American and male is a significant predictor of probation failure. Additionally, being a Hispanic felon was also statistically significant. Theoretical, practical, and research implications are discussed alongside study limitations.