This article presents results from a study that evaluated an online message system designed to improve the delivery of prevention programs. We conducted a quasi-experimental study with 32 agencies and schools that implemented substance use prevention programs and examined differences between the comparison and intervention groups. We also examined the impact of dosage of the message system by comparing results among three groups of teachers: non-users, low users, and high users. Results for norm setting were marginally significant, such that teachers within the agencies assigned to the intervention condition scored higher on their understanding of norm setting at posttest compared to teachers within comparison agencies, after controlling for pretest knowledge scores and demographic items. In the model examining impact of dosage, high users of the intervention scored significantly higher on self-reported understanding of their program, quality of delivery, and program effectiveness compared to non-users. Low users of the intervention reported significantly higher quality of delivery compared to non-users.