The school environment has been often identified as a prosperous venue for public health improvement. This study is a cluster randomized controlled trial evaluating the impact of a school-based helmet promotion program on knowledge, attitudes and practices of eligible adolescent drivers. Four public, four private and four vocational high schools situated in Attica, Greece, were sorted by type and randomly assigned to receive a 1-month intervention, based on the concepts of the Health Belief Model, or serve as controls. Self-report data were collected at baseline from 741 second grade students (16 years) and immediately after program completion. Linear mixed models with random student effects were used to estimate mean changes in scores for each treatment group and corresponding between groups differences of changes. Likelihood-based analysis showed that the intervention yielded a significant improvement in knowledge about helmet use. Yet, its impact on attitudes and practices appeared to vary across different school types. With current research offering ambiguous results on the appropriate timing of injury prevention efforts, this study suggests that educational programs targeting road safety can lead to positive changes if tailored to the needs of specific population groups and implemented during critical life periods, such as the transition to driving status.