This study sought to evaluate how the addition of a general practitioner (GP) surgery influences the utilization of an emergency department (ED). An intervention trial with historical control was conducted in a Swedish university hospital ED. A GP surgery was established in the ED by the addition of GP physicians without the addition of other personnel (nurses, secretaries, aids). The number of persons evaluated and managed by the GP physicians and ED physicians were quantified preintervention (April 1992 to October 1993) and postintervention (April 1994 to October 1995). Further information was obtained by questionnaires distributed to all physicians and patients during three sample study weeks: 1 week before intervention and 6 and 18 months after the intervention. Patient volume, percentages of inappropriate visits, and types of services were recorded. The addition of GP physicians increased the number of visits to the ED by 27% (4,694 per month to 5,952 per month). The percentage of patients managed in the ED who had nonurgent complaints (primary health care needs) increased with the intervention from 22% (95% confidence interval [CI] 19%, 25%) to 33% (95% CI 30%, 37%). The increased demand on the ED of patients with nonurgent complaints increased the average waiting time for patients with urgent or emergent complaints from 35 minutes to 40 minutes (14%). The introduction of GPs to an ED increased the number and proportion of patients presenting to the ED with nonurgent complaints.