Stable isotope analysis of bone provides insight into animal foraging and allows for ecological reconstructions over time, however pre-treatment is required to isolate collagen. Pre-treatments typically consist of demineralization to remove inorganic components and/or lipid extraction to remove fats, but these protocols can differentially affect stable carbon (δ13C) and nitrogen (δ15N) isotope values depending on the chemicals, tissues, and/or species involved. Species-specific methodologies create a standard for comparability across studies and enhance understanding of collagen isolation from modern cetacean bone. Elemental analyzers coupled to isotope ratio mass spectrometers were used to measure the δ13C and δ15N values of powdered killer whale (Orcinus orca) bone that was intact (control) or subjected to one of three experimental conditions: demineralized, lipid-extracted, and both demineralized and lipid-extracted. Additionally, C:N ratios were evaluated as a proxy for collagen purity. Lastly, correlations were examined between control C:N ratios vs. historical age and control C:N ratios vs. sample characteristics. No significant differences in the δ15N values were observed for any of the experimental protocols. However, the δ13C values were significantly increased by all three experimental protocols: demineralization, lipid extraction, and both treatments combined. The most influential protocol was both demineralization and lipid extraction. Measures of the C:N ratios were also significantly lowered by demineralization and both treatments combined, indicating the material was closer to pure collagen after the treatments. Collagen purity as indicated via C:N ratio was not correlated with historical age nor sample characteristics. If only the δ15N values from killer whale bone are of interest for analysis, no pre-treatment seems necessary. If the δ13C values are of interest, samples should be both demineralized and lipid-extracted. As historical age and specimen characteristics are not correlated with sample contamination, all samples can be treated equally.