During a 21-month period, 1223 gynaecological patients were studied consecutively during 3 hours in the recovery room. Three primary anaesthetics were randomly utilized: (a) cyclopropane, with or without thiopentone; (b) halothane, with or without thiopentone; and (c) thiopentone-nitrous oxide-oxygen. Ten per cent of the patients received miscellaneous agents. The overall incidence of emesis was approximately 29 per cent. Cyclopropane was associated with a significantly higher incidence of emesis than were halothane and thiopentone-nitrous oxide-oxygen. Thiopentone induction followed by a primary inhalation agent was associated with a lower incidence of emesis than was induction with the primary agent alone. The duration of anaesthesia, the lithotomy in contrast to the supine position, the type of premedication and the specific type of gynaecological surgery were not significantly influential nor were many miscellaneous physical or pharmacological factors. The chief conclusion to be drawn from this study is that the primary anaesthetic agent and the persence or absence of a thiopentone induction remained the most vital influences on the incidence of post-anaesthetic emesis in the recovery room.