The microscopic method of age at death determination was introduced by Kerley in 1965. The method, which relies on the quantification of selected elements in cortical bone tissue, has been widely used, and several other researchers have modified or added to the method. Yet, very few studies have been carried out dealing with the intra- and inter-observer error. Furthermore, when such studies have been completed, the statistical tools for assessing variability have not been adequate. This study presents the results of applying simple quantitative statistics on several counts of microscopic elements as observed on photographic images of cortical bone, in order to assess intra- and inter-observer error. Overall, substantial error was present at the level of identifying and counting secondary osteons, osteon fragments and Haversian canals. Only secondary osteons can be reliably identified, precluding the use of osteon fragments and Haversian canals. The observers in this study included experienced and inexperienced users of the microscopic method, yet the variability was uniformly large for all observers, suggesting fundamental problems in definition and identification of the structural elements. Until more rigorous definitions of such elements have been agreed upon, the use of microscopical methods must be discouraged as a sole or uncontrolled method of evaluating age at death.