Ninety patients undergoing scheduled upper limb orthopaedic surgery were studied to determine the optimal anaesthetic dose using the ‘orthogonal two-needle technique’. The patients were randomly assigned to one of three groups to receive one of three different volumes (20, 30 and 40mL) (n = 30) of anaesthetic solution (a mixture of equal parts of 0.5% bupivacaine with adrenaline 1:200 000 and 2% lignocaine). A significant correlation was found between the volume injected and the anaesthetic spread for all tested areas. A better analgesic spread to all the major branches of the plexus was obtained when increased volumes of anaesthetic solution were injected. The comparisons between the 20 mL group and the other two groups are significant in all the tested areas, as well as the comparisons between 30 and 40 mL groups in the areas innervated by radial and musculo-cutaneous nerves. Only the area innervated by the axillary nerve showed a weaker volume-analgesia relation, confirming the elusiveness of this area to anaesthesia in the axillary approaches. The improved results observed using greater amounts of anaesthetic solution might result from a higher intrasheath pressure with disruption of sheath septa, or from a greater availability of drug for all the terminal branches of brachial plexus, or both.