Cyanide poisoning is uncommon, but generates interest because of the presumed utility of an antidote immediately available in those areas with a high risk of cyanide exposure. As part of its regular review of guidelines, the Australian Resuscitation Council conducted a systematic review of the human evidence for the use of various proposed cyanide antidotes, and a narrative review of the relevant pharmacological and animal studies. There have been no relevant comparative or placebo-controlled human trials. Nine case series were identified. Treatment with hydroxocobalamin was reported in a total of 361 cases. No serious adverse effects of hydroxocobalamin were reported, and many patients with otherwise presumably fatal poisoning survived. Sodium thiosulphate use was reported in two case series, similarly with no adverse effects. Treatment with a combination of sodium nitrite, amyl nitrite and sodium thiosulphate was reported in 74 patients, with results indistinguishable from those of hydroxocobalamin and sodium thiosulphate. No case series using dicobalt edetate or 4-dimethylaminophenol were identified, but successful use in single cases has been reported. Hydroxocobalamin and sodium thiosulphate differ from alternatives in having negligible adverse effects, and on the basis of current evidence are the antidotes of choice. The indications for the use of an antidote, the requirements for supportive care and a recommended approach for workplaces where there is a risk of cyanide poisoning are presented.