Although diazepam has been shown to reduce the stress response, the protective effect of midazolam, a newer benzodiazepine from a stressful event, tracheal intubation, has not been studied as yet by catecholamine assays in patients undergoing coronary artery bypass surgery, who also receive intravenous sufentanil as a component of the neuroleptanalgesic technique. Therefore, we evaluated the influence of midazolam in combination with sufentanil on the plasma free catecholamines before and after midazolam, after sufentanil and pancuronium and before and after intubation in 15 adult patients undergoing coronary artery surgery. After routine premedication, midazolam 0.14 +/- 0.01 mg.kg-1 i.v. was given over 1 min followed 5 min later by sufentanil in incremental i.v. doses of 1.5 micrograms.kg-1 to a total pre-intubation dose of 4.0-5.0 micrograms.kg-1 injected in 10 min. The incremental doses of sufentanil were given when a greater than 15 per cent increase in rate-pressure product occurred. One min after the initial dose of sufentanil, pancuronium 0.1 mg.kg-1 i.v. was given to provide muscle relaxation. Midazolam administration per se caused a significant decrease in systolic and diastolic blood pressures with a concomitant reduction in systemic vascular resistance. Sufentanil reduced the left ventricular stroke-work index. Tracheal intubation, a strong stressor during anesthesia, elicited no increase in catecholamines and/or adverse hemodynamic responses in contrast to a marked increase in plasma catecholamines routinely observed in patients anesthetized by the commonly used technique of intravenous barbiturates in combination with succinylcholine.