BACKGROUND Levobupivacaine is the S(-)enantiomer of bupivacaine, a long acting amino-ester local anaesthetic agent. Cocktail mixture of levobupivacaine and fentanyl infusion is commonly given via epidural for post-operative analgesia. The concentration of levobupivacaine for optimal pain relief with least side effects remained uncertain. METHODS This was a prospective, randomized double-blind study to evaluate the effectiveness and safety of two different concentrations of levobupivacaine as epidural infusion for postoperative pain relief. Fifty patients who underwent orthopaedic lower limb surgeries were recruited and given a standardized combined spinal epidural anaesthesia during operation. They were then allocated to receive either 0.1% (Group A) or 0.2% levobupivacaine (Group B) with 2 Mi g/ml fentanyl as epidural infusion for postoperative pain relief over 24 hours. Pain score, motor blockade, hemodynamic parameters and the need for rescue analgesia were recorded. RESULTS Group B patients had significant lower pain score at 4 and 8 hours post operation. There was no significant difference in degree of motor blockade and need for rescue analgesia in both groups. However, patients from Group B had significant hypotension (24% vs 4%) that responded to volume resuscitation without life-threatening complications. CONCLUSIONS 0.2% levobupivacaine with 2 Mi g/ml fentanyl provided more superior analgesia compared to 0.1% levobupivacaine with 2 Mi g/ ml fentanyl in patients who underwent lower limb orthopaedic surgery, but with a significant higher incidence of hypotension that responded to volume replacement.