Halting further spread of multidrug-resistant tuberculosis (MDR-TB) requires both new resources and a renewed discussion of priority setting informed by estimates of the existing burden of this disease. The 1997 report of the first phase of the global survey by the World Health Organization (WHO) and the International Union Against Tuberculosis and Lung Disease (IUATLD) uses the indicator of the proportion of TB cases that are MDR-TB to identify MDR-TB 'hot spots'. We sought to refine the definition of MDR-TB transmission 'hot spots'. For this purpose, we obtained estimates of two additional indicators for regions where data are available: MDR-TB incidence per 100 000 population per year, and expected numbers of new patients with MDRTB per year. There is generally much agreement in the three indicators considered, and some differences also appear. We conclude that it is useful, when defining indicators of MDR-TB transmission 'hot spots', to include estimates of underlying TB incidence rates and of absolute numbers of MDR-TB cases. Estimating the force of morbidity of MDR-TB in a population is important for comparing this burden across settings with very different underlying TB incidence rates: estimating the absolute number of MDR-TB patients will be critical for planning the delivery of directly observed MDR- TB therapy and the rational procurement of second-line drugs. Through this exercise, we aim to initiate discussion about improved methods of quantifying and comparing current MDR-TB transmission 'hot spots' that require intervention.