Background and objective Insertion of the laryngeal mask airway in the anaesthetized patient can sometimes be difficult and propofol has been advocated as the anaesthetic induction agent of choice because of its depressant effect on laryngeal reflexes compared with other intravenous anaesthetics. However, when used as the sole induction agent, relatively large doses of propofol are required to achieve successful laryngeal mask insertion. This has cost implications and may produce unwanted cardiorespiratory depression.

Methods One hundred and forty-two patients were randomized to receive either: fentanyl 1 μg kg−1 and lidocaine 1.5mgkg−1 (group 1), or fentanyl 1μgkg−1 and midazolam 0.04 mg kg−1 (group 2), or fentanyl 1μg kg−1, midazolam 0.04 mg kg−1 and lidocaine 1.5mgkg−1 (group 3) or fentanyl 1 μg kg−1 (group 4) 2min before induction of anaesthesia. Anaesthesia was established with propofol infused at 33.3mgmin−1.

Results Patients who were given midazolam required significantly less propofol to achieve satisfactory laryngeal mask insertion, median propofol doses: group 1, 1.63mgkg−1; group 2, 1.16mgkg−1; group 3, 1.01 mgkg−1; group 4, 1.9mgkg−1, P < 0.0001 (analysis of variance). Patients given midazolam reported less pain on injection with propofol 13% and 3% groups (2 and 3) compared with 37.5% and 77% (groups 1 and 4) P = 0.002 (χ2).

Conclusions Midazolam reduces the dose of propofol required for induction of anaesthesia and successful insertion of the laryngeal mask airway. There was no clinical benefit to be gained from the addition of lidocaine.