PURPOSE: To determine what kind of violence-related behavior or opinion is directly related to excessive media use among adolescents in Switzerland. METHODS: A national representative sample of 4222 schoolchildren (7th- and 8th-graders; mean age 13.9 years) answered questions on the frequency of television-viewing, electronic game-playing, feeling unsafe at school, bullying others, hitting others, and fighting with others, as part of the Health Behaviour in School-Aged Children (HBSC) international collaborative study protocol. The Chi-square tests and multiple logistic regression analyses were applied to high-risk groups of adolescents. RESULTS: For the total sample, all bivariate relationships between television-viewing/electronic game-playing and each violence-related variable are significant. In the multivariate comparison, physical violence among boys ceases to be significant. For girls, only television-viewing is linked to indirect violence. Against the hypothesis, females' electronic game-playing only had a bearing on hitting others. CONCLUSIONS: Experimental designs are needed that take into account gender, different forms of media, and violence to answer the question of whether excessive media use leads to violent behavior. With the exception of excessive electronic game-playing among girls, this study found that electronic media are not thought to lead directly to real-life violence but to hostility and indirect violence.