School-based programs seem promising for child abuse prevention. However, research mainly focused on sexual child abuse and knowledge is lacking on how individual program components contribute to the effectiveness of school-based prevention programs for any form of child abuse. This study aimed to examine the overall effect of these school-based programs on (a) children’s child abuse-related knowledge and (b) self-protection skills by conducting two three-level meta-analyses. Furthermore, moderator analyses were performed to identify how program components and delivery techniques were associated with effectiveness. A literature search yielded 34 studies (158 effect sizes; N = 11,798) examining knowledge of child abuse and 22 studies (99 effect sizes; N = 7804) examining self-protection skills. A significant overall effect was found of school-based programs on both knowledge (d = 0.572, 95% CI [0.408, 0.737], p < 0.001) and self-protection skills (d = 0.528, 95% CI [0.262, 0.794], p < 0.001). The results of the first meta-analysis on children’s child abuse knowledge suggest that program effects were larger in programs addressing social–emotional skills of children (d = 0.909 for programs with this component versus d = 0.489 for programs without this component) and self-blame (d = 0.776 versus d = 0.412), and when puppets (d = 1.096 versus d = 0.500) and games or quizzes (d = 0.966 versus d = 0.494) were used. The second meta-analysis on children’s self-protections skills revealed that no individual components or techniques were associated with increased effectiveness. Several other study and program characteristics did moderate the overall effects and are discussed. In general, school-based prevention programs show positive effects on both knowledge and self-protection skills, and the results imply that program effectiveness can be improved by implementing specific components and techniques.