The study evaluated the impact of rotational assignment of emergency department (ED) patients to residents on patient's length of stay (LOS) and resident satisfaction. The study was conducted in a university, inner-city, adult ED. Prior to the intervention, residents saw patients at their own rate as patient charts were placed into a common rack waiting to be seen. The intervention involved directly assigning patients with medical emergencies in sequential rotation to residents. Patients with surgical conditions were seen under the old self-paced system for comparison. A retrospective chart review of LOS of all ED visits for a period of 2 wks before, 6 wks after, and 1 yr after the intervention was conducted. ED administration and staffing during the study were unchanged. Comparisons excluded critical cases and nonurgent cases triaged away. A postintervention survey of physician satisfaction was conducted. During the study period, demographic characteristics of medical and surgical patients did not differ significantly between the time periods. The average total ED LOS for moderately ill medical patients decreased significantly from 7.11 to 5.86 h at 6 wks and remained significantly improved 1 yr later at 6.21 h. During the same periods, the average total LOS increased significantly for the surgical patients. Residents reported that the new system was more fair and did not affect teaching quality. The rotational assignment of patients to resident physicians led to significantly less ED LOS while improving resident satisfaction without affecting education.