This article aims to present a simple methodology to estimate the coverage of registration of deaths and obtain estimates of mortality for small areas, using the example of Minas Gerais. The proposed methodology combines the structure of indirect standardization of mortality of smaller areas based on the functions of mortality of larger areas and then applying the methods of distribution of deaths for more robust estimates of mortality for smaller areas. The results are robust and suitable compared to government estimates, and using other methods. The estimates produced using the two stages procedure, were robust and reduced cases of over- registration of deaths counts and improved estimates of adult mortality in some areas where data are less reliable.
mortality; underregistration; demography