Among the therapeutic approaches available to manage canine nasal tumors, radiation therapy is considered the standard of care for achieving improved quality of life and overall survival time for these patients. Long-term local tumor control remains the goal for many radiation therapy protocols applied to patients with canine nasal tumors. Stereotactic radiation therapy (SRT), utilized by Colorado State University Animal Cancer Center (CSU-ACC), delivers maximum dose to the tumor volume, while preferentially sparing the surrounding normal structures. When treating dogs with nasal SRT, there is uncertainty with regard to the actual dose delivered to the oral mucosa due to the anatomical proximity of the tumor and the inherent error of most algorithms in calculating doses at air-tissue interfaces. A canine head phantom, dogs with spontaneously-occurring tumors, and the Gafchromic EBT film analysis system were utilized to measure radiation dose. Further comparison between measured film dose and dose calculated by the Varian Eclipse inverse planning algorithm was conducted. The results from this study demonstrate the high variability of the Analytical Anisotropic Algorithm (AAA) in calculating doses for this particular geometry. The high degree of variability results in uncertainty with regard to the prediction of delivered dose. The differences between measured and calculated doses in both the phantom (mean difference of 104 cGy per fraction, p-value of <0.0001) and the dogs (mean difference of 74 cGy per fraction, p-value of 4.2 x 10^-5), suggest that the AAA underestimates the dose delivered to oral mucosa of dogs treated with SRT to manage their nasal tumors. Based upon a dose prescription of 10 Gy tumor dose per fraction in this population of dogs, an oral mucosa dose range of 3.5-6.5 Gy per fraction was measured in 88% of the cases, this range might be considered by clinicians in an effort to estimate the dose delivered to the oral mucosa while treating canine nasal tumors with SRT.