Porotic hyperostosis is currently considered to be one of several stress markers available for assessing the health and nutritional status of past human populations. The present study questions one of the basic assumptions underlying its use; that is, that the occurrence of porotic hyperostosis in an individual represents an episode of anemia that was current or had occurred within a relatively short period prior to death. A synthesis of data from a Romano-British site Poundbury Camp, anthropological and clinical studies, and information on bone physiology suggests that lesions of porotic hyperostosis seen in adults are most probably representative of a childhood episode of anemia. Lesions seen in adults are the result of bone changes occurring in the growth period that have not undergone complete remodelling. This viewpoint has implications for future interpretation of data on porotic hyperostosis obtained from skeletal collections.