The brachial plexus in infraclavicular region can be blocked by various approaches. Aim of this study was to compare two approaches (coracoid and clavicular) regarding success rate, discomfort during performance of block, tourniquet tolerance and complications. The study was randomised, prospective and observer blinded. Sixty adult patients of both sexes of ASA status 1 and 2 requiring orthopaedic surgery below mid-humerus were randomly assigned to receive nerve stimulator guided infraclavicular brachial plexus block either by lateral coracoid approach (group L, n = 30) or medial clavicular approach (group M, n = 30) with 25-30 ml of 0.5% bupivacaine. Sensory block in the distribution of five main nerves distal to elbow, motor block (Grade 1-4), discomfort during performance of block and tourniquet pain were recorded by a blinded observer. Clinical success of block was defined as the block sufficient to perform the surgery without any supplementation. All the five nerves distal to elbow were blocked in 77 and 67% patients in groups L and M respectively. Successful block was observed in 87 and 73% patients in groups L and M, respectively (P > 0.05). More patients had moderate to severe discomfort during performance of block due to positioning of limb in group M (14 vs. 8 in groups M and L). Tourniquet was well tolerated in most patients with successful block in both groups. No serious complication was observed. Both the approaches were equivalent regarding success rate, tourniquet tolerance and safety. Coracoid approach seemed better as positioning of operative limb was less painful, coracoids process was easy to locate and the technique was easy to learn and master.