Multiple‐choice (MC) anatomy “spot‐tests” (identification‐based assessments on tagged cadaveric specimens) offer a practical alternative to traditional free‐response (FR) spot‐tests. Conversion of the two spot‐tests in an upper limb musculoskeletal anatomy unit of study from FR to a novel MC format, where one of five tagged structures on a specimen was the answer to each question, provided a unique opportunity to assess the comparative validity and reliability of FR‐ and MC‐formatted spot‐tests and the impact on student performance following the change of test format to MC. Three successive year cohorts of health science students (n = 1,442) were each assessed by spot‐tests formatted as FR (first cohort) or MC (following two cohorts). Comparative question difficulty was assessed independently by three examiners. There were more higher‐order cognitive skill questions and more of the course objectives tested in the MC‐formatted tests. Spot‐test reliability was maintained with Cronbach’s alpha reliability coefficients ≥ 0.80 and 80% of the MC items of high quality (having point‐biserial correlation coefficients > 0.25). These results also demonstrated guessing was not an issue. The mean final score for the MC‐formatted cohorts increased by 4.9%, but did not change for the final theory examination that was common to all three cohorts. Subgroup analysis revealed that the greatest change in spot‐test marks was for the lower‐performing students. In conclusion, our results indicate spot‐tests formatted as MC are suitable alternatives to FR tests. The increase in mean scores for the MC‐formatted spot‐tests was attributed to the lower demand of the MC format.