We analyzed carbon stable isotope data from bone collagen of animals consuming varied experimental diets, including recently published data from Warinner and Tuross [Warinner, C., Tuross, N., 2009. Alkaline cooking and stable isotope tissue-diet spacing in swine: archaeological implications. Journal of Archaeological Science 36, 1690-1697; this journal]. Comparing regression lines for the relationship between collagen and diet δ13C, we show that protein source, and not physiology, explains the apparent taxonomic difference between swine and rodents reported in that paper. Our results reveal a complex relationship between whole diet and dietary protein in determining collagen δ13C values, such that in many cases, collagen alone may not provide reliable reconstructions of paleodiet. We advocate the simultaneous use of both collagen and apatite δ13C, whenever possible, to assess the diets of prehistoric peoples. © 2010 Elsevier Ltd.