Anatomy education is an inherently visual field, particularly in bellringer (BR) testing, which requires learners to identify anatomical structures on human-donated specimens. While the traditional use of these physical specimens in BR testing has long remained the standard, three-dimensional (3D) viewing in virtual reality platforms and two-dimensional (2D) images of specimens on paper have become common alternatives due to the ease and feasibility of use. Despite widespread use, there is a paucity of literature comparing the assessment validity of these modalities to physical specimens as the historic standard. Thus, this study sought to assess BR testing performance and question validity (using point biserial evaluation) across all three modalities. In total, 140 undergraduate students, enrolled in an Introductory Anatomy and Physiology course at the time of testing, participated in a BR examination with specimens presented in three visual formats: physical specimens, printed 2D images, and 3D reconstructions in virtual reality. When comparing all three modalities, no notable differences were found between question difficulty, point biserial values, presentation of cybersickness, visuospatial ability, or modality preference. Additionally, modality preference and student opinion did not significantly affect test scores, suggesting that these student attributes were unrelated to BR performance. The examinations had high reliability as measured by KR-20 values, supporting the applicability of our results to undergraduate anatomy BR testing. This study provides preliminary evidence supporting the utility and validity of both 2D images and 3D virtual reality as alternative modalities for BR testing within the undergraduate anatomy education setting.