BACKGROUND: The position for tracheal intubation using direct laryngoscopy is extension of the head with flexion of the neck, the classical 'sniffing position'. If necessary, an extra pillow can be used to keep the neck flexed. By adopting this position the oral, pharyngeal, and laryngeal axes is a almost straight line to facilitate tracheal intubation. Also, this position is ideal for conventional laryngeal mask airway (LMA) insertion. However, insertion of intubating laryngeal mask airway (ILM) and intubation through ILM may be achieved from any position relative to the patient's head. As recommended by the manufacturer, when possible a pillow should be placed under the head to achieve a neutral position. The purpose of this study was therefore to compare the easiness of intubation through ILM without support and with the patient's head supported by a pillow. METHODS: After acquiring informed consent, 80 ASA grade 1 or 2 patients undergoing general anesthesia for elective surgical procedures who normally required tracheal intubation were randomized into two groups. In group 1 (n = 40), insertion of ILM and intubation was conducted with the head supported by a pillow, while there was no support in group 2 (n = 40). The patients were induced and relaxed with an IV injection of thiopental sodium, fentanyl-ketamine-midazolam mixture and vecuronium. When adequate level of anesthesia was achieved, the ILM was inserted. After adequate ventilation was confirmed, a blind tracheal intubation through the ILM was attempted. Then we recorded success rate, insertion time, intubation time and adjusting maneuvers. RESULTS: The ILM was successfully inserted on the first attempt in 79/80 patients, but 1 patient of group 1 failed to be adequately ventilated. The mean time for ILM insertion of group 2 was shorter than that of group 1. The success rate of tracheal intubation was 37(95%) in group 1 and 40 (100%) in group 2. In group 1, 30 (81%) patients were successfully intubated on the first attempt, 1 (3%) patient on the second attempt, and 6(16%) patients on the third attempt; in group 2, 35 (87%) patients on the first attempt, and 5 (13%) patients on the third attempt. There was no significant diffrence of mean time taken for endotracheal intubation through ILM between group 1 (105.1 sec) and group 2 (88.1 sec). CONCLUSIONS: The authors conclude that ILM insertion is significantly easier with the patient's head not supported by a pillow compared with the patient's head supported by a pillow and there is no difference in ease of intubation through ILM by the patient's head position.