We compared the effectiveness of atropine, glycopyrrolate, and a transesophageal atrial pacing (TAP) stethoscope for treating intraoperative bradycardia in 64 unpremedicated patients receiving a standardized sufentanil/NO2/vecuronium anesthetic. Patients were allocated randomly to receive either atropine, 5 μg/kg (Group 1), glycopyrrolate, 2.5 μg/kg (Group 2), or transesophageal atrial pacing (Group 3) after the onset of bradycardia, defined as a heart rate of ≤50 beats/ min (or ≤60 beats/min with concurrent hypotension). Bradycardia occurred in 15 patients of each treatment group. The time required for the heart rate to increase to ≥70 beats/min was 270 (range 30-490), 270 (70-465), and 12 (2-30) s in Groups 1, 2, and 3, respectively. Although all patients in Group 3 responded to pacing at 150% of the threshold current, 10 patients in Group 1 and 8 patients in Group 2 required a second dose of anticholinergic medication before a heart rate response was observed. One patient in Group 2 required three doses, and another who did not respond even after four doses was treated with the TAP device. Bradycardia subsequently recurred in five patients in Group 1 and four patients in Group 2. Temporary recurrence of bradycardia occurred in seven patients in Group 3 due to outward migration of the pacing stethoscope. However, heart rates were more consistently maintained in paced patients. There were no significant differences in postoperative side effects between the three groups, or when compared with patients who did not receive treatment for bradycardia. In conclusion, TAP resulted in more rapid and reliable treatment of intraoperative bradycardia than either atropine or glycopyrrolate, but did not result in a decreased incidence of side effects.