The time, material, and staff-consuming nature of anatomy's traditional pen-and-paper assessment system, the increase in the number of students enrolling in medical schools and the ever-escalating workload of academic staff have made the use of computer-based assessment (CBA) an attractive proposition. To understand the impact of such shift in the assessment method, an experimental study evaluating its effect on students' performance was designed. Additionally, students' opinions toward CBA were gathered. Second-year medical students attending a Clinical Anatomy course were randomized by clusters in two groups. The pen-and-paper group attended two sessions, each consisting of a traditional sectional anatomy steeplechase followed by a theoretical examination, while the computer group was involved in two similar sessions conducted in a computerized environment. At the end of each of the computer sessions, students in this group filled an anonymous questionnaire. In the first session, pen-and-paper group students scored significantly better than computer-group students in both the steeplechase (mean ± standard deviation: 66.00 ± 14.15% vs. 43.50 ± 19.10%; P < 0.001) and the theoretical examination (52.50 ± 12.70% vs. 39.00 ± 21.10%; P < 0.001). In the second session, no statistically significant differences were found for both the steeplechase (59.50 ± 17.30% vs. 54.50 ± 17.00%; P = 0.085) and the theoretical examination (57.50 ± 13.70% vs. 54.00 ± 14.30%; P = 0.161). Besides, an intersession improvement in students' perceptions toward CBA was registered. These results suggest that, after a familiarization period, CBA might be a performance equivalent and student accepted alternative to clinical anatomy pen-and-paper theoretical and practical examinations. Anat Sci Educ 11: 124-136. © 2017 American Association of Anatomists.