Three clinical trials are reported in which diazepam was compared with other premedications hitherto used for ENT operations under local anaesthesia. The first trial, in 60 tonsillectomy patients, compared diazepam (15 mg) with Thalamonal (a mixture of droperidol and fentanyl) 1 ml. Apart from a larger fall in blood pressure caused by Thalamonal no major differences between both premedications were found. The second trial in 66 patients undergoing nasal septum correction compared diazepam (20 mg) with a complex routine involving the use of triflupromazine, barbitone, promethazine and pethidine. This mixture caused dryness of the mouth and drowsiness more often than did diazepam, and in some cases also respiratory depression. In the third trial 125 patients undergoing endoscopy received either diazepam 15 mg or placebo. A significantly higher proportion of the patients who received diazepam were relaxed. Taking the three trials together, minor and serious complications were less frequent in patients pretreated with diazepam than in those who received premedications hitherto in use. Intramuscular diazepam alone thus offers a satisfactory alternative premedication for these procedures.