A frequently encountered task in the forensic scenario is verification of the human origin of severely degraded fragments of bone. In these cases histological methods which consider osteon size and morphology can prove to be useful. The authors in the present study verify the applicability of published algorithms to flat and subadult bones from human, dog, cat, cow, rabbit, sheep, pig, chicken, quail, and turkey samples. Metric analysis was performed on 2031 Haversian canals. Analyses carried out on human samples confirmed a success rate of around 70% on long adult bones; however the percentage of wrong answers was particularly high in the case of newborns and older subadults as well as on flat bones in general. Results therefore suggest that such regression equations should be limited only to bone fragments from long adult bones.