Anaesthesia is to be congratulated for publishing the Difficult Airway Society (DAS) Guidelines (Anaesthesia 2004; 59: 675–94), which I am sure will result in a reduction in anaesthesia-related morbidity and mortality. However, I was amazed to read that rejection was considered, presumably on the basis of concerns raised in the accompanying editorial by Chambers (Anaesthesia 2004; 59: 631–3). Comment has been encouraged. Chambers challenges the assertion that failed intubation is the leading cause of anaesthetic mortality and suggests that the DAS guidelines may actually cause more problems than they solve. I agree that contemporaneous data is lacking and that this would be extremely difficult to obtain for such a rare event. All the more reason therefore to rely on expert opinion. In my hospital, over the last 5 years, there have been two anaesthesia-related deaths, one in theatre and one in the Intensive Care Unit (ITU), both due to failed intubation. This is a very approximate incidence of 1 in 100 000 general anaesthetics and 1 in 1000 ITU admissions. In my experience therefore failed intubation is the leading cause of anaesthesia-related mortality, entirely justifying a robust approach to prevention. Chambers seems particularly concerned about the place of surgical cricothyroidotomy. The guidelines make it absolutely clear that this is the last resort, is only to be used in the ‘can’t intubate, can't ventilate' (CICV) situation and that cannula cricothyroidotomy with high-pressure ventilation is an equally valid alternative in plan D. ‘Very difficult ventilation’ is not an indication per se and trainees are not ‘encouraged to proceed to cricothyroidotomy’ in this situation alone. It is precisely because junior trainees, or even senior consultants, may not be able to differentiate between ‘impossible’ and ‘only just sufficient’ ventilation that the emphasis is on the prevention of hypoxia, which can be more objectively recognised. There is no place for a definitive level of peripheral blood oxygen saturation (Spo2) at which cricothyroidotomy is indicated, as the risk of hypoxic damage will be relative to the rate of desaturation and individual patient resistance. This requires clinical judgement, but if there is doubt, I welcome a guideline that gives a definitive solution to a potentially lethal problem. In my hospital, it is most unlikely that the first of our two cases could have been salvaged by cricothyroidotomy, because of a massive neck haematoma. By the time tracheostomy was achieved, the patient was dead. The ITU case had been intentionally extubated but could not be re-intubated at the first attempt. He might have been salvaged by cricothyroidotomy, but the patient suffered an early cardiac arrest due to his underlying condition and a transiently unrecordable Spo2, despite some ventilation being possible with a bag and mask. I have successfully performed a surgical cricothyroidotomy in a patient with extensive facial trauma, increasing oedema and impending airway obstruction. I, too, have over 20 years experience in anaesthesia and, up to that point, had never experienced a CICV situation. It all went wrong very quickly, despite careful thought and preparation. I am convinced that the patient is alive and well today because I had trained myself for this eventuality. This is why DAS believes that cricothyroidotomy of one type or another should be a core skill for all anaesthetists. The modest chance of resolvable morbidity that Chambers is so concerned about is likely to be far outweighed by the high risk of unresolvable hypoxic brain damage or mortality. A patient should never die from failed intubation, but only after failed cricothyroidotomy.