(Midazolam is often used for premedication; it is known to promote vasodilation and may therefore affect redistribution of heat during surgery. We examined the effect of pre-operative administration of midazolam on the development of intra-operative hypothermia. Forty-five patients were randomly allocated to one of three groups to receive no premedication (Group C), IM midazolam 0.04 mg.kg(-1) (Group M1) or 0.08 mg.kg(-1) (Group M2) 30 min prior to anaesthesia. Sedation levels were assessed, and then general anaesthesia was induced and maintained using propofol and fentanyl. During surgery, core temperature, which was similar for the three groups prior to induction of anaesthesia, decreased significantly less in the midazolam groups M1 and M2 compared to the control group C. Patients who were more heavily sedated prior to induction of anaesthesia, had significantly lower core temperatures peri-operatively than those who were less sedated, and core temperatures in unpremedicated patients fell to significantly lower levels during surgery than those who were drowsy. We conclude that pre-operative administration of midazolam produces an effect on the development of peri-operative hypothermia. We found that moderate pre-operative sedation reduces the peri-operative heat loss, possibly by affecting core-to-peripheral heat distribution.