Abstract RATIONALE Stable isotope analysis is a frequently used tool in ecology for unraveling dietary dynamics and trophic relationships. In isotopic studies that use mixing models to infer diet, a single tissue is often used to represent the isotopic values of a whole organism. Here, we examine that assumption in keratin-based tissues and provide recommendations for isotopic diet studies of animals that consume prey whole. METHODS By comparing carbon ( δ ¹³C) and nitrogen ( δ ¹⁵N) stable isotope values in single tissues of mammals and birds, this study aims to elucidate the extent to which nondestructive sampling methods in tissues such as hair and feathers, accurately represent the overall isotopic composition of four mammalian species (rats, mice, rabbits, guinea pigs) and one avian species (quail). RESULTS We found that keratin-based tissue and the homogenized whole-body samples were significantly different in δ ¹³C values in mammals (mean 𝚫 ¹³C -1.93 ± 0.37‰) but not in birds. Feathers differed significantly from whole body in δ ¹⁵N values in birds (mean 𝚫 ¹⁵N -0.32 ± 0.11‰), but δ ¹⁵N values did not differ among hair and the whole body of mammals. CONCLUSIONS We provide a mathematical correction for diet reconstruction based on specific tissues, especially for animals consuming whole prey. We suggest that future studies reconsider the assumption that non-invasively sampled tissues are representative of whole-body isotopic values.