Caudal bupivacaine is widely used for inguinal hernia repair in children, but often cannot totally eliminate responses to hernial sac traction. The current study examined whether supplementation of caudal bupivacaine with dexmedetomidine could achieve better results. Sixty children aged 1272 months undergoing unilateral inguinal hernia repair received standardized premedication with midazolam, i.v. ketamine anaesthesia, and then were randomly assigned to receive either bupivacaine 0.25 (1 ml kg(1); Group B) or bupivacaine plus dexmedetomidine (1 g kg(1); Group BD). The response to hernial sac traction was defined as an increase in heart rate or systolic arterial pressure by 20, and was treated with ketamine rescue (2 mg kg(1)). After the surgery, fentanyl was administered as needed with a nurse-controlled analgesia pump. Only one subject in Group BD (3.33) needed ketamine rescue, as opposed to 13 subjects in Group B (43.33; P0.001). The first fentanyl injection occurred at a much later time point in Group BD (median: 860 vs 320 min in Group B; P0.001). Total fentanyl consumption of fentanyl was significantly lower in Group BD [2.5 (1.2) vs 6.9 (1.6) g kg(1) 24 h(1) in Group B; P0.008]. The addition of dexmedetomidine to caudal bupivacaine could reduce the response to hernial sac traction, and prolong the duration of postoperative analgesia in children undergoing inguinal hernia repair.