We have studied 40 healthy children, aged 3-10 yr, undergoing adenotonsillectomy, in a double-blind, randomized study. Intubating conditions were assessed when the pupils had become small and central after inhalation induction with either 5% halothane and 60% nitrous oxide in oxygen or 8% sevoflurane and 60% nitrous oxide in oxygen. The quality of tracheal intubation was graded according to ease of laryngoscopy, position of the vocal cords, coughing, jaw relaxation and movement of limbs. Fewer children had significant vocal cord movement on laryngoscopy (P < 0.01) and more had ideal intubating conditions when halothane was used (12 of 20 compared with seven of 20; ns). Time to reach the clinical end-point for intubation was reached sooner with halothane (P = 0.015). In all children the trachea was intubated successfully at the first attempt and all remained haemodynamically stable throughout induction.