This paper considers some of the procedures used and problems encountered in the collection of demographic data for a large-scale medical project: the Endemic Diseases Research Project at Malumfashi in northern Nigeria, 1974-1979. 43,216 people were enumerated and a subset of 26,100 were visited monthly for one year for the purpose of registration of vital events. The majority of problems encountered are relevant to similar studies in developing countries. Some difficulties cannot be avoided, but it is considered helpful to identify them. Other situations, however, deserve particular attention, such as the development of adequate on-site data-checking mechanisms, as these are areas in which future programmes could benefit from the experience of the demographic studies in Malumfashi. More effort still needs to be put into the improvement of data collection methods themselves.