The aim of this experiment was to develop an alternative, acceptable and humane method for killing broiler chickens. Our main premise was that euthanasia should preferably take place in two steps. Firstly, the bird should be anaesthetised quickly and completely using appropriate procedures and, secondly, the anaesthetised bird should be killed as quickly as possible. Two gas mixtures were considered for anaesthesia: 40% carbon dioxide with 30% oxygen and 30% nitrogen or 40% carbon dioxide with 15% oxygen and 45% nitrogen. It was found that the birds' behaviour in the high oxygen environment was less disturbed, with less agitation and distress being demonstrated. For this reason, the high oxygen condition is to be preferred and this anaesthetic was used in the second part of the experiment. The bird was exposed for one minute to 40% carbon dioxide, 30% oxygen and 30% nitrogen (anaesthesia) and then to 80% carbon dioxide, 5% oxygen and 15% nitrogen (euthanasia). This condition was compared with the frequently used eutha- nasia condition involving argon. The use of argon is based on the idea that it produces fast suffocation of the birds. The argon condition consisted of 30% carbon dioxide and 60% argon in air with less than 2% oxygen and 8% nitrogen. The chickens died significantly later in the oxygen/carbon dioxide condition than in the argon condition. Judgements of death were based on both the pattern of electrical brain activity and on heart rate. The time taken for the chickens to become unconscious, as judged from the isolectric electroencepha- logram, was also significantly longer in the oxygen/carbon dioxide condition than in the argon condition. However, in the argon condition the birds immediately showed strong signs of agitation and distress, particularly intense and heavy muscle contractions, whereas these signs were less pronounced in the oxygen/carbon dioxide condition. It is concluded that, although the chickens lost consciousness sooner and also died sooner in the argon condition, there are good reasons to prefer the oxygedcarbon dioxide condition because it results in less disturbed behaviour and the birds exhibited milder signs of distress during the period when consciousness cannot be fully excluded. Hence, a milder death, taking somewhat longer, is preferred to a quicker but more distressing death. Finally, the oxygen/carbon dioxide condition was repeated with turkeys when, in general, the results were the same.