Oral diazepam 0.25 mg/kg or 0.5 mg/kg was employed as premedication in one hundred and one children undergoing elective surgery. The drug failed to modify the rise in cardiorespiratory indices of preanaesthetic anxiety compared with control values, and there was no difference between the two doses assessed by a sedation scoring system. 0.25 mg/kg diazepam produced less sedation in children under 5 years old, compared with those 5 years and older, whereas 0.5 mg/kg produced no difference between the older and younger age groups. The plasma levels of diazepam were greater postoperatively in the 0.5 mg/kg group. There was no relationship between plasma diazepam and recall at induction, and pre-anaesthetic amnesia was not enhanced with the higher premedication dose.