A re-evaluation of dilute (8 per cent w/v) ethanol as a sole intravenous anaesthetic agent (with various premedicants), in 300 patients undergoing minor gynaecological surgery, is reported. It confirms that adequate doses will rapidly induce sleep. Sedative or analgesic premedication, notably with promethazine, can make ethanol a more effective and acceptable induction agent, with the maximum dosage (44 g) employed in this study but it was not as reliable for induction as the barbiturates or eugenol derivatives. Its safety in experienced hands was demonstrated and cardiovascular stability was a notable feature. Marked individual variation in response was observed widi this agent, most notably after chlordiazepoxide premedication, where a significantly higher dosage was required to produce sleep than with control (atropine) patients. Though complications were rare during anaesthesia, the incidence of emergence delirium, headache and emetic sequelae with this agent were higher than with the barbiturates or eugenols. Although these could be minimized by heavy opiate premedication, the consequent respiratory depression and prolonged drowsiness made this undesirable. This study has made it possible to observe the various interactions of ethanol with premedicant drugs, and an antagonism by large doses of chlordiazepoxide has been demonstrated.