In this article we review the evidence supporting the clinical application of adrenaline in cardiopulmonary arrest, and summarize the receptor effects of catecholamines and the basic principles producing perfusion during CPR. Animal and human studies show that in cardiac arrest, adrenaline has positive haemodynamic effects, increasing systemic pressures, myocardial perfusion, and cerebrally directed flow. The problems extrapolating from animal to human data are highlighted. Studies showing improvements in short term survival outcomes with high dose regimens have not been confirmed by other large prospective randomised trials. There is no evidence that high doses of adrenaline improve survival to hospital discharge. Most studies comparing adrenaline with placebo have been non-randomised and uncontrolled, with major methodological problems. Conclusions are difficult, but if anything adrenaline is associated with poorer outcomes.