One hundred and twenty-seven children aged 3-6 years were allocated to four groups. All of them received venous cannulation on the dorsum of the hand. On induction, the group L1, L2 and L3 patients received propofol 3 mg/kg mixed with lignocaine 0.15 mg/kg, 0.3 mg/kg, 0.6 mg/kg, respectively. The group T patients received thiopentone 3 mg/kg, then propofol 1.5 mg/kg mixed with lignocaine 0.075 mg/kg. Pain on injection was categorized into two-assessment items (facial expression and limbs withdrawal). The facial expression category were subdivided into none, mild (knit of brows), moderate (grimace), and severe (crying). The withdrawal of limbs was categorized into none, mild (withdrawal of hand), moderate (withdrawal of fore-arm and arm), severe (withdrawal of arm and twisting of body). Patients were monitored using an electrocardiogram, pulse oximeter, autonomic noninvasive blood pressure measuring device and capnography. The patient characteristics did not differ significantly among the four groups. Pain on injection was significantly more frequent in the group L1 patients (81%) compared with the group T (27%) patients. Increasing lignocaine dose reduced the incidence of pain graded as "moderate" or "severe" though there was no significant difference. The incidences of excitatory effect on propofol injection were reduced with increasing lignocaine dose and prior administration of thiopentone but there were no obviously differences among groups. We concluded that thiopentone reduced injection pain on propofol and should be recommended.