Social network interventions targeted at children and adolescents can have a substantial effect on their health behaviours, including physical activity. However, designing successful social network interventions is a considerable research challenge. For example, it is unclear which criteria should be used to select influence agents that serve as successful promoters of the targeted health behaviour. Investigating this question through field experiments is a time consuming and infeasible process. Fortunately, advancements in computer science enable us to simulate these complex processes. In this work, we rely on social network analysis and agent-based simulations in order to better understand and capitalize on the complex interplay of social networks and health behaviours. More specifically, we investigate which criteria for selecting influence agents can be expected to produce the most successful social network health interventions.The aim is to use a computational model to simulate different selection criteria for social network interventions and to observe the intervention’s effect on the physical activity of primary and secondary school children within their school classes. As a next step, this study will rely on the outcomes of the simulated interventions to investigate whether social network interventions are more effective in some classes thanothers based on network characteristics.We used a previously validated agent-based model to understand how physical activity spreads in social networks and who is influencing the spread of behaviour. Based on the observed data of 460 participants collected in 26 school classes, we simulated multiple social network interventions ranging in selection criteria for the influence agents (i.e. in-degree, betweenness and closeness centrality and random influence agents) and a control condition (i.e. no intervention condition). Subsequently, we investigated whether the detected variation of an intervention’s success within school classes could be explained by structural characteristics of the social networks (i.e. network density and network centralization).The one-year simulations showed that the social network interventions were more effective compared to the control condition, β = .30, t(100) = 3.23, P = .001. In addition, the social network interventions that used a measure of centrality to select influence agents outperformed the random influence agent intervention, β = .46, t(100) = 3.86, P < .001. Also, the closeness centrality condition outperformed the betweenness centrality condition, β = .59, t(100) = 2.02, P = .046. The anticipated interaction effects of the network characteristics were not observed. Social network interventions can be considered as a viable and promising intervention method to promote physical activity. We demonstrated the usefulness of applying social network analysis and agent-based modelling as part of the social network interventions’ design process. We emphasize the importance of selecting the most successful influence agents and provide a better understanding of the role of network characteristics on the effectiveness of social network interventions.