This paper provides an evaluation of an experiment in using discrete event simulation modelling at a Rapid Improvement Event in a large hospital trust. It presents empirical findings about the challenges of building a model rapidly within a time constrained event. The aim was also to help introduce Lean principles so the learning and understanding acquired by participants, their interaction with the model, and the experimentation promoted and facilitated by the model are all considered. Our learning from this action research about rapid modelling and the modifications made to our approach as a result of the experiment are described.