The objective of this study was to document that an observational study decreases the use of the emergency department (ED) for asthma. Comparison of rates between an audit and prospective period were used at a regional referral pediatric hospital ED. A total of 526 asthma visits from February 12, 1992, to April 10, 1992, were examined in an initial audit and compared to 725 visits during a prospective study from May 16, 1993, to September 29, 1993. A physician check list for medications and follow-up plans was utilized during the prospective study. The rate of repeat visits and admissions was compared between the audit and the prospective periods. In the audit, 422 asthma patients were seen, and there were 68 repeat visits within one week of the initial visit and 153 admissions. For 29 of the 68 repeat visits there was no documentation that medication had been increased after the initial visit. No follow-up plans were recorded in 275 of the 526 visits. During the prospective period 668 children presented to the ED with asthma, and 346 were enrolled into the study on risks for repeat visits. There were 57 repeat visits and 89 admissions. The repeat visit rate during the prospective period was 9 and 8%, respectively, in the enrolled and nonenrolled subjects. The repeat visit and admission rate decreased during the prospective period as compared to during the audit by 39% (P = 0.004) and 58% (P < 0.0001), respectively. A possible explanation for the difference in admission and repeat visit rates between the audit and prospective study is that physicians' behavior may have been altered by the awareness that their behavior was being closely monitored during the study.