Background and Objectives: Burnout among residents can have numerous impacts on training and career plans. Accurate estimates of burnout are crucial, and multiple specialties have assessed resident burnout via surveys conducted after an In-Training Exam (ITE). Our objective was to assess difference in self-reported symptoms of burnout by survey context. Methods: In 2025, we used 2024 American Board of Family Medicine data from final year residents who, in overlapping time periods, answered two items on burnout after the ITE and during registration for the Initial Certification Examination. We compared changes in responses on each instrument by bivariate tests with resident characteristics. Adjusted regression analyses determined characteristics associated with reporting burnout after the ITE but not during exam registration. Results: The rate of reported burnout was higher after the ITE than when registering for the certification exam (49.9% vs 30.4%). More residents reported burnout after the ITE but not when registering for the certification exam (897, 22.5%) than the opposite (119, 3.0%). Residents' ITE performance was not associated with reported burnout differences. Conclusions: Residents had significantly higher odds of reporting burnout when assessed after an ITE than when registering for a certification examination in an overlapping time period. These findings suggest that burnout estimates are highly sensitive to survey context, even when measured in the same residents over an overlapping time period.