In developed countries, the post-mortem examination may not be critical to the identification of patients with cancer, since diagnostic sophistication and histological verification are readily available. The autopsy should be considered of increasing importance, however, to epidemiologists and other investigators who use the data generated by cancer registries. The contribution of post-mortem examinations to etiological studies, to documentation of the natural history of disease in a changing environment and to evaluating diagnostic and therapeutic modalities should, in fact, be recognized as essential to the practice of modern medicine and to research. In the less developed countries, cancer registration must depend to a much greater degree on necropsy for identification of cancer patients, since surgical biopsy may be limited or selected and often specimens are not even submitted for examination. The primary site and histology of a tumour are frequently not available from clinical records, which may carry significant inaccuracies. Most of the important observations and contributions relating to the geographical pathology of cancer have made substantial use of post-mortem material. All of the investigative approaches that are dependent on reliable cancer registry data and that involve use of information derived from autopsies are equally applicable to developed and less developed countries. Such information is of enormous importance to public health planning, and the potential for disease prevention cannot be overestimated.