Human cadaveric prosections are a traditional, effective, and highly appreciated modality of anatomy learning. Plastic models are an alternative teaching modality, though few studies examine their effectiveness in learning of upper limb musculoskeletal anatomy. The purpose of this study is to investigate which modality is associated with a better outcome, as assessed by students' performance on examinations. Overall, 60 undergraduate medical students without previous knowledge of anatomy participated in the study. Students were assigned into two groups. Group 1 attended lectures and studied from cadaveric prosections (n = 30) and Group 2 attended lectures and used plastic models in the laboratory (n = 30). A knowledge assessment, including examination with tag questions (spot test) and written multiple-choice questions, was held after the end of the study. Students' perceptions were also investigated via an anonymous questionnaire. No significant difference in students' performance was observed between the group using prosections and the group using plastic models (32.2 ± 14.7 vs 35.0 ± 14.8, respectively; P = 0.477). Similarly, no statistically significant difference was found regarding students' satisfaction from using each learning modality (P = 0.441). Plastic models may be a valuable supplementary modality in learning upper limb musculoskeletal anatomy, despite their limitations. Easy to use and with no need for maintaining facilities, they are highly appreciated by students and can be useful when preparing for the use of cadaveric specimens.