The reported incidence of amputation varies enormously between countries, races and communities, but gives an indication of the suffering and costs caused by disease of the foot in diabetes. However, the field is highly complex and interpretation of available data requires great care; many factors underlie the differences which have been reported. Community-wide information is not widely available, and most published work derives from selected populations. Published work is also characterised by lack of consensus in the use of terms such as "lower extremity amputation", "major" and "minor" amputation, as well as by the use of inconsistent criteria for the selection of operations for study. Some publications refer to all operations performed, while others refer to the latest undertaken on any one limb or in any one patient. Incidence may be expressed in terms of either the total (diabetic and non-diabetic), or the "at risk" (diabetic) populations. The former may be useful as a (partial) measure of disease burden in a community, while the latter can be used as an indirect measure of the quality of overall management of diabetes in a population. However, the extent to which either can be used as a measure of the quality of specialist care of an individual presenting with limb-threatening disease is uncertain. Ultimately, an amputation is only a treatment, and in no other medical condition is the number of operations, or other treatments, used as an indication of either disease burden or the quality of care. Quality of care of foot disease in diabetes can, and should, be best assessed in terms of survival, function/incapacity and wellbeing.