Emergency medical dispatch (EMD) is typically a team activity, requiring fluid coordination and communication between team members. Such working situations have often been described in terms of distributed cognition (DC), a framework for understanding team working. DC takes account of factors such as shared representations and artefacts to support reasoning about team working. Although the language of DC has been developed over several years, little attention has been paid to developing a methodology or reusable representation which supports reasoning about an interactive system from a DC perspective. We present a case study in which we developed a method for constructing a DC account of team working in the domain of EMD, focusing on the use of the method for describing an existing EMD work system, identifying sources of weakness in that system, and reasoning about the likely consequences of redesign of the system. The resulting DC descriptions have yielded new insights into the design of EMD work and of tools to support that work within a large EMD centre.