Purpose: The Bispectral Index (BIS) has been recently shown to objectively predict the level of sedation in patients undergoing conscious sedation. It was the goal of this study to directly compare the recovery profile of patients where the BIS was used to monitor sedation with a control group where the monitor was not used. Patients and Methods: Forty patients undergoing third molar extractions under intravenous conscious sedation were randomly assigned to 2 groups. In both groups, induction of sedation was performed using a standard dose of fentanyl (1.5/μg/kg) and midazolam (0.05 mg/kg). Propofol was then given in 10 to 20 mg boluses until a clinically desirable sedation level was achieved. In 1 group, the BIS was then monitored continually during surgery using a microcomputer (Aspect-1050 Monitor, Aspect Co, Natick, MA) and recorded at 5-minute intervals. The anesthetist (N.A.S.) provided additional propofol boluses to maintain a BIS level of 70 to 80. In the other group, the BIS sensor was applied, but the monitor was not used. In this group, the sedation was modified, and additional propofol was given based solely on the anesthetist's subjective assessment of the desired level of sedation (Observer's Assessment of Alertness/Sedation [OAA/S] scale level 2 to 3). Additional boluses of 1 mg of midazolam were given during the procedure if patients required repeated boluses of propofol at less than 5-minute intervals to maintain the desired sedation level (BIS level of 70 to 80 or OAA/S level of 2 to 3). These additional midazolam boluses, as well as the time of the last sedative dose (propofol or midazolam) were recorded to study the effect of these factors on recovery. Results: Of the 40 patients initially included in the study, 1 subject in the BIS-monitored group was excluded due to the loss of intravenous access at initiation of the case. For the remaining 39 subjects, 19 were assessed objectively using the BIS monitor, whereas 20 were assessed subjectively using the OAA/S scale. The BIS cases were slightly longer in duration than the OAA/S cases, lasting an average of 26 minutes versus 22 minutes. This difference was statistically nonsignificant (P = . 19). Less propofol was used in the BIS cases, with an average of 98 mg for BIS cases versus 106 mg for OAA/S cases (P = .59). The total dose in mg/kg/min was significantly less in the BIS group (0.054 mg/kg/min) than in the OAA/S group (0.074 mg/kg/min; P = .0082). There was no significant difference in the amount of midazolam administered after induction between the 2 groups (P = .60). The surgeon, who was blinded to whether the monitor was used, ranked the third molar extractions more difficult in the BIS group (P = .05). However, patients in the BIS group were on average more cooperative, with better maintenance of muscle tone. The difference in these parameters were nonsignificant (P = .15 and .092, respectively). A positive Romberg test was obtained earlier in BIS patients, although this difference was nonsignificant (P = .097). The straight-line test was completed significantly sooner in BIS patients (P = .013). There was no significant difference between the BIS and OAA/S groups in perceptual speed (P = .55) or computation (P = .32). There was essentially no difference between groups in patient-assessed comfort or recall of the procedure. There were also no notable differences in anesthesia complications, return to activities of daily life, or pain medication use between the 2 groups. Conclusions: The BIS provides additional information for standard monitoring techniques that helps guide the administration of sedative-hypnotic agents. It appears that use of the BIS monitor can help to titrate the level of sedation so that less drugs are used to maintain the desired level. The trend toward an earlier return of motor function in BIS-monitored patients warrants further investigation.