Type 2 diabetes is globally increasing in prevalence and is widely recognized as a major cause of morbidity and mortality, as well as being a burden to the health-care services. Planning for current and future diabetes services requires up-to-date prevalence information. The enumeration of Type 2 diabetes is, however, surprisingly difficult. Large numbers of people are undiagnosed, and those known cases have variable loci of care. Traditional techniques include cross-sectional diagnostic surveys, postal or house-to-house surveys and cohort surveys. All are time-consuming and expensive, and may potentially undercount. The use of multiple patient lists (e.g. hospital clinic data, general practitioner (GP) lists, prescribing information, etc.) can, however, increase accuracy and, if the data are computerized, may be rapid and inexpensive. A new and potentially exciting tool to utilize multiple lists in Type 2 diabetes prevalence assessment is known as 'capture-recapture'. In this, statistical models are used to estimate prevalence from the degree of overlap between lists. Capture-recapture is emerging as a valuable tool in the epidemiological assessment of Type 2 diabetes.