Objective:To compare the application effects of high-definition two-dimensional (HD-2D) and glasses-free three-dimensional (GF-3D) display systems in thoracoscopy teaching.Methods:A total of 40 clinical medicine interns with no surgical experience from The First Affiliated Hospital of Guangzhou Medical University were recruited and were required to participate in a 1-week training course of endoscopy. They were then randomly allocated to the HD-2D group and GF-3D group and asked to perform three tasks: peg transfer, circular cutting, and suture knotting. Their performance was measured with a system that scored speed and precision. SPSS 25.0 was used to conduct t-test, Pearson Chi-square test and Fisher exact test for the comparison. Results:The mean time for the peg transfer test in GF-3D group was shorter than that in HD-2D group, without statistically significant difference [(63.20±21.11) s vs. (71.15± 17.26) s, P = 0.212]. The mean time for the circular cutting test in GF-3D group was shorter than that in HD-2D group, without statistically significant difference [(112.50±16.67) s vs. (118.15±24.43) s, P=0.410]. The mean time for the suture knotting test in GF-3D group was shorter than that in HD-2D group, with statistically significant difference [(301.50±32.77) s vs. (341.75±57.23) s, P=0.019]. The total score in GF-3D group was higher than that in HD-2D group, with statistically significant difference [(78.33±5.88) points vs. (72.08±6.83) points, P=0.005]. Conclusion:The GF-3D display system is clearly superior to the HD-2D system because it reduces the surgical learning curve, and is therefore suitable for basic teaching and skills training.