We have assessed prospectively the time to readiness for surgery following axillary block (sum of block performance and latency times) in 80 patients. The brachial plexus was identified using a nerve stimulator, and anaesthetized with 45 mL of mepivacaine 1% with adrenaline 5 micrograms mL-1. In group 1 (single injection) the whole volume of mepivacaine was injected after locating only one of the plexus nerves. In group 2 (multiple injections) at least three plexus nerves were located, and the volume of mepivacaine was divided between them. Sensory block was assessed by a blinded observer every 10 min. Patchy analgesia was supplemented after electrolocating the unblocked nerves after 20, 30 or 40 min. The patient was pronounced ready for surgery when analgesia was present in all areas to be operated upon, which always included the three nerves to the hand. The single injection technique required less time for block performance (mean 5.5 min) than multiple injections (mean 9.5 min), P < 0.0001. However, latency of the block was longer and the requirement for supplemental nerve blocks was greater, after single injections (33 min and 57%) than after multiple injections (15.5 min and 7%, respectively), P < 0.0001. As a result, readiness for surgery was achieved faster in group 2 (25 min), than in group 1 (38.5 min), P < 0.0001. After supplementation, block effectiveness was 100% in group 1 and 98% in group 2 (NS). The frequency of adverse effects (vessel puncture or paraesthesia) was similar in both groups. No neurological sequelae were observed. We conclude that the multiple injection technique takes longer to perform than single injection, but that readiness for surgery is faster because of shorter block latency and better spread of analgesia.