I read with interest the letter by Ravi Shankar, Stacey and Ravalia (Anaesthesia 1999; 54: 1023) regarding limitations of the current Checklist for Anaesthetic Machines and wish to add my personal concerns At no stage in the current Association of Anaesthetists guidelines [1] is the audible oxygen failure alarm activated, as the pipeline supply remains connected throughout the test. This contrasts with the previous checklist [2] in which (with the pipeline disconnected) the oxygen cylinder, after checking its contents, was turned off resulting in the primary audible alarm sounding as the oxygen pressure decreased. This familiar sound was instantly recognised by all anaesthetists, and many other operating theatre staff, who would hear it at least once before each list commenced. Having adopted the current Checklist for Anaesthetic Apparatus since its introduction in 1997, I was taken by surprise when, during a recent case, our oxygen failed and the audible alarm sounded. It took me a couple of seconds to recognise its significance and thus correct the problem. My concern is that we may breed a generation of anaesthetists for whom that once familiar audible warning of oxygen failure is unrecognised and its significance unknown. Thank you for the opportunity to reply to Dr Taylor. As Chairman of the Working Party that revised the checklist, I can assure him that the decision to omit the pipeline disconnection, and hence checking of the oxygen failure alarm, was not taken lightly. It was discussed at length in the Working Party, and in the end omitted on the grounds that pipeline system failure is exceedingly rare: at the time of re-writing the check list there had been no reported failures of which we or the Medical Devices Agency were aware. There was, however, evidence that anaesthetists were not checking machines before use, with the then existing check list, because it was too complex. One of the difficult steps was that of pipeline disconnection and re-connection, and we decided that a checklist that people would use was preferable to one that was widely ignored. The checklist, as published, is simple and easily completed and is intended for the anaesthetist to ensure that his/her equipment is safe to use. Further, fuller checks are perfectly in order by ODAs etc. during which the oxygen failure device can be checked. Checking of the oxygen failure device is mandatory if using cylinders as the only source of oxygen. Perhaps the next generation of anaesthetists will have less need of audible warnings of oxygen failure, as they will rarely, if ever, work from cylinders only. Dr Taylor raises an important point, but if this is the first occasion in 2 years in which he has had cause to hear the oxygen alarm, I rest my case. The case which Dr Shankar reported is also of some concern, but the facts remain: pipeline failure is exceedingly rare. D. P. Cartwright Derby City General Hospital, Derby DE22 3NE, UK