School psychologists and other school professionals are increasingly turning to digital citizenship programs to improve children's online safety. However, to date, there has been no rigorous outcome evaluation of these programs. The current study used a cluster randomized control trial methodology (CRCT) to evaluate the impact of the Be Internet Awesome digital citizenship program on knowledge, attitudinal, and behavioral outcomes for 1072 4th-6th graders at 14 participating schools (7 treatment and 7 control schools). Multilevel logistic, multinomial, and Gaussian models were used to test outcomes. The study found support for program impact on children's knowledge of new online safety concepts and self-efficacy for handling online problems relative to students in control conditions. However, no program impact was found for other outcomes including online privacy or civility, cyberbullying, and talking with parents about online problems. Findings highlight the need for clarity on digital citizenship program goals, innovative approaches, and increased evaluation research.