OBJECTIVES: Computer simulation studies of the emergency department (ED) often allow the patient to drive the process and do not consider indirect patient related activities by the attending physician and resident (i.e. charting, teaching). The objective of this study is to describe and evaluate an approach where physicians are considered ‘pseudo-agents’ in a discrete event simulation (DES). METHODS: Using data from an Ontario hospital, two ED DES models (traditional versus pseudo-agent) were constructed and compared on key outputs including time intervals and utilization rates. The traditional approach models the attending physician and residents as human resources where their only function is to treat patients. Additionally, residents are modeled similarly to the attending. The pseudo-agent approach allows for interaction between the attending physician and residents, the use of different skill sets (attending treats more severe patients) and decision-making hierarchy (physician prioritizes on treating a patient or teaching a resident). These concepts were implemented using the simulation software ARENA 13. RESULTS: Utilization of nursing staff and clerks remained similar. Physician utilization rates increased by 50% in the pseudo-agent based model. There was no change in the time from when the patient was placed in a room until the first physician visit. Time until discharge or admission increased for patients with a lower acuity in the pseudo-agent based model. As a result, time to room for these patients also increased in addition to bed utilization.  CONCLUSIONS: Computer simulation models are more frequently being used to help inform hospital operational decision making. This example shows the importance of accurately modeling physician relationships and the rules in which they treat patients. Neglecting these relationships could lead to inefficient resource allocation due to inaccurate estimates of resource utilization and waiting times.