Propofol is one of the most frequently applied intravenous anesthetics for the induction of general anesthesia. However, pain on injection of this agent is a considerable problem in daily anesthesia practice because of its severity. Administration of lidocaine prior to propofol injection is a standard technique for reducing the pain on injection. However, this method provides insufficient pain relief. To evaluate whether pretreatment with an ultra-short acting barbiturate, thiopental, is more effective than with lidocaine, a randomized and single-blinded trial was conducted. Patients (20-65 years old, n=137) were allocated into six groups, and applied with physiological saline, thiopental (25, 50, 75, or 100mg), or lidocaine (40mg) at 30 second prior to propofol injection (1mg/kg, 1200ml/h). The patient was interviewed about the degree of pain just after propofol was totally injected. Both thiopental (≧25mg) and lidocaine decreased the severity of pain in comparison with physiological saline as evaluated by a six-graded pain score. Lidocaine failed to influence the incidence of pain (from 86% to 55%), although thiopental significantly decreased it to 40% (25mg), 21% (50mg), 12% (50mg), and 0% (100mg), respectively. Thiopental (≧50mg) decreased both the severity and incidence of pain more effectively than lidocaine. A Hill plot analysis of these data, after rearrangement by patient's body weight, estimated that the half-effective dose (ED_<50>) and the ED_<99> of this drug to block pain on injection of propofol were 0.6 and 1.4mg/kg, respectively.