The delivery device attached to any medical laser system is the very important, if somewhat neglec- ted, "Cinderella" of the system. It is the part that the surgeon or operator interarts with and yet de- spite occupying this key position there are few, if any, objective standards to which the delivery de- vices have to conform. There is a requirement, for example, that if the power output from a CO2 laser varies outside the limits set at + 20 % a system should at least indicate a fault. Despite providing some indication of the output power calibrated in such a way as to represent the output power of the end of the articulating arm there is no requirement on continuous monitoring at the end of the arm and therefore, should a mirror fail through being contaminated with smoke that variation would not be recorded nor any action taken unless the drop in power was noted by the surgeon. I am not necessarily argueing in favour of distal power monitoring but simply make the point that failures of the delivery system have not yet been incorporated into formal regulations.