OBJECTIVE To review the evidence that buffering of local anaesthetics with sodium bicarbonate reduces the pain of injection whilst not affecting efficacy. METHODS Medline search from 1966 to December 2001. Articles in all languages were included. Bibliographies were examined for papers. RESULTS The search identified 63 publications. All were retrieved. Of these, 22 were human prospective randomized controlled trials directly assessing the pain of infiltration. Three papers were based on observations. No case series, case reports, or retrospective studies were identified. One animal study was found. CONCLUSIONS The evidence is that buffering with sodium bicarbonate significantly reduces the pain of local anaesthetic injection. The buffered solutions retain the efficacy of local anaesthetics and are stable in the mixtures used in the trials. Adrenaline-containing buffered solutions need refrigeration in closed containers for storage. Buffering will be particularly useful where pain of local anaesthetic injection may not be well tolerated such as in large areas of infiltration, sensitive areas such as the face and in children. It is recommended that sodium bicarbonate and tables of stable dilutions are readily available in the emergency department to facilitate this.