Intravenous (IV) propofol was compared with IV thiopental/pentobarbital as a sedative for children undergoing magnetic resonance imaging (MRI) of the brain or spine. Fifty-eight outpatients (aged 11 mo to 6 1/2 yr, ASA grade I and II) were enrolled in the study and randomized to two groups. After IV cannulation, Group I received IV propofol (1-2 mg/kg), followed immediately by a propofol infusion (75-100 micrograms.kg-1.min-1). Group II received IV thiopental (1-3 mg/kg) followed by a pentobarbital bolus (2-3 mg/kg). Supplemental thiopental doses (1-2 mg/kg) were administrated to maintain adequate sedation. Discharge time and postanesthesia recovery scores were determined by an independent blinded observer. Time of recovery to full consciousness in Group I was significantly less than in Group II (19 +/- 7 min vs 35 +/- 20; P < 0.005). Time to discharge was also significantly less in Group I (24 +/- 6 min vs 40 +/- 11; P < 0.05). A preliminary cost analysis was applied to the clinical data obtained and to a theoretical model of a pediatric MRI center. Cost analysis of anesthesia services revealed added drug costs ($1600.76 per year for the propofol group) but significant savings of postanesthesia care unit (PACU) nursing time ($5086.67 per year). Outcomes such as patient morbidity and technical quality of the MRI scans did not differ significantly between the two groups. In conclusion, analysis of the clinical data suggests that propofol may be more suitable than barbiturates for children undergoing outpatient procedures despite its higher price.