This paper describes how palaeopathological or historical data can be used to investigate the comparative importance of internal or external causes of diseases that may affect bone. Two examples of the method are given: osteoarthritis of the costo-vertebral and costo-transverse joints of the spine, and the infectious disease, osteomyelitis. In the former case we suggest that the palaeopathological data support the view that internal factors are much the most important, since the characteristics of the disease are similar in skeletons widely separated in time and space. In the latter case, in which historical medical data have been used, it seems probable that the decrease in the prevalence of the disease in recent times has been due primarlly to external factors, particularly improvements in hygiene and diet. We hope that these findings will stimulate others to apply this methodology to other pathological considtions affecting the skeleton.