A method for objective evaluation of the difficulty of endotracheal intubation is described. Our data indicate that the angle formed by the light-beam axis of the laryngoscope blade and the laryngotracheal axis, which we call "angle phi", is analogous to the degree of difficulty of endotracheal intubation. Using this method, we compared the effectiveness of a standard Macintosh and a modified bevelled Macintosh blade in 27 tracheostomized Intensive Care Unit patients under general anaesthesia. Statistical analysis of our results indicate that the bevelled blade significantly facilitates endotracheal intubation.