Measurement of the collapsibility of the upper airway while a patient is awake is not a good guide to such collapsibility during sleep, presumably because of differences in respiratory drive, muscle tone, and sensitivity of reflexes. To assess whether a relation existed between general anaesthesia and sleep, we measured collapsibility of the upper airway during general anaesthesia and severity of sleep-disordered breathing in 25 people who were having minor surgery on their limbs. Anaesthetised patients who needed positive pressure to maintain airway patency had more severe sleep-disordered breathing than did those whose airways remained patent at or below atmospheric pressure. Such an association was strongest during rapid-eye-movement (REM) sleep. Our findings suggest that sleep-disordered breathing should be considered in all patients with a pronounced tendency for upper airway obstruction during anaesthesia or during recovery from it.