High-dose narcotic anesthetic inductions usually avoid circulatory depression better than do other techniques; however, the selection of a narcotic and neuromuscular blocker influences subsequent hemodynamic responses. One hundred-one patients having aortocoronary bypass graft (CABG) surgery were investigated using four combinations of a narcotic and neuromuscular blocker: group FP (fentanyl 50 micrograms/kg, pancuronium 100 micrograms/kg); group FV (fentanyl 50 micrograms/kg, vecuronium 80 micrograms/kg); group SP (sufentanil 10 micrograms/kg, pancuronium 100 micrograms/kg); and group SV (sufentanil 10 micrograms/kg, vecuronium 80 micrograms/kg), each combination being administered over 2 minutes. Hemodynamic functions were then monitored for 10 minutes before tracheal intubation. Significant changes included increases in heart rate in the groups receiving pancuronium and decreases in those receiving vecuronium. In all groups mean arterial pressure initially decreased; systemic vascular resistance index decreased significantly in all groups except SV. Cardiac index decreased significantly only in group SV. Circulatory depression requiring treatment with vasopressor or anticholinergic drugs was more common in patients given vecuronium. Cardiac arrhythmia occurred most often in group SP; only in group FP were there no arrhythmias, ischemic changes, or hemodynamic disturbances requiring intervention. Time to onset of neuromuscular blockade did not differ among the four groups, but transient chest wall rigidity occurred significantly more often with sufentanil than with fentanyl. Overall, the fentanyl/pancuronium combination afforded the greatest hemodynamic stability, whereas the sufentanil/vecuronium combination proved least satisfactory because of bradycardia and hypotension, requiring treatment in 35% of group SV patients.(ABSTRACT TRUNCATED AT 250 WORDS)