The objective of the research presented in the dissertation at hand is to improve situation awareness in tactical mobile ad hoc networks (MANETs). Tactical teams are supported to successfully accomplish their missions. This could be a team of first responders after a natural catastrophe or a terrorist attack. In such scenarios mobile devices connected by a MANET are a quick, flexible and efficient way to provide a backup communication infrastructure. We provide three main contributions in order to increase the robustness of situation awareness in tactical MANETs: cross data analysis, cooperative trust assessment and probabilistic state modeling. Cross data analysis provides a framework for exploiting all available data sources (direct and indirect sensor data, mission-specific knowledge and general information sources) in order to detect inconsistencies. The focus of cooperative trust assessment is to identify nodes with wrong or malicious behavior and exclude them from the situation awareness process. The probabilistic state modeling and estimation concept based on particle filters allows incorporating multiple information sources at several stages to adjust the likelihood of specific system states.