Vessel Traffic Service (VTS) is a shore-side maritime assistance service that supports bridge teams in their safe navigation of port approaches and other areas that present navigational difficulties. The VTS is implemented in national waters and provides vessels with information through transmissions and broadcasts on Very High Frequency (VHF) radio. With a continued growth in the number, size and cargo volumes of merchant vessels, the role of the VTS has recently become a matter of discussion, and it has been argued that changes, such as implementing an aviation-like control system, would be of an enormous benefit for stakeholders and guarantee safe and efficient traffic movements in the future. The complexity of processes in safety-critical domains, such as maritime traffic management, is increasing due to continuing technical, organisational and environmental developments. The VTS is currently undergoing drastic changes, primarily driven by strategies and projects focusing on increasing the overall efficiency of the maritime transportation system through advanced technology. To reduce the risk of unforeseen consequences, it is important to study and understand the service and its contribution to traffic management before changes are implemented. The purpose of this thesis has been to increase the overall understanding of everyday performance of the VTS system and identify ways of modelling the performance of the service, as a contribution to the ongoing debate on the future needs of maritime traffic management. The VTS is described as socio-technical system that controls and manages maritime traffic in port approaches and other areas that pose navigational difficulties for bridge teams. Field data collected through semi-structured interviews, observations and focus groups have been analysed with the aid of concepts derived from Cognitive Systems Engineering (CSE) and Resilience Engineering (RE) to understand how the VTS actively contributes to safety through monitoring, responding to and anticipating changes in traffic patterns in the VTS area. The data have also been used to model performance variability in everyday operation with the aid of the Functional Resonance Analysis Method (FRAM). Performance variability is necessary for a system to be adaptive, and is therefore essential for the system’s functioning. By using the FRAM, a new angle of the VTS system has been explored to understand how variability in its functional units affects the overall system performance. The thesis demonstrates the importance of understanding how performance in a socio-technical system can vary and the consequences this may have. The FRAM can be used to analyse the functional design of a socio-technical system, and therefore help to identify and assess ways in which performance variability can be monitored and managed. By understanding the functional design of the VTS system and the complexity of everyday operation, stakeholders will be able to identify advantages and disadvantages of current system design and use this to consider how future demands can best be met.