The purpose of this study was to determine whether low-fidelity arthroscopic simulation training improves basic ankle arthroscopy performance and efficiency among orthopedic trainees. Twenty-nine orthopedic surgery trainees with varying levels of experience in ankle arthroscopy were randomized into either simulation or standard practice groups. At baseline testing, all participants performed simulator-based testing and a cadaveric diagnostic ankle arthroscopy with video recording. The simulation group subsequently received 4 one-on-one, 15-minute simulation training sessions over a 4-month period, while the standard practice group received no additional simulation training or exposure. After intervention, both groups were reevaluated with simulator testing and a second recorded cadaveric diagnostic ankle arthroscopy. Two blinded, independent experts evaluated each randomized arthroscopic performance using the 15-point checklist, Arthroscopic Surgery Skill Evaluation Tool (ASSET), and total elapsed time, and all outcome measures were compared within and between groups. Baseline arthroscopic experience, simulator task performance measures, and ASSET scores were equivalent between the simulation and standard practice groups. After completion of training, the simulation group outscored the control group in total ASSET score (34.9 vs 19.6; P < .001) and checklist score (14.5 vs 8.4; P < .001) and achieved nearly expert ASSET Safety scores (4.7 vs 2.9; P < .001) on the simulator model. Cadaver testing also demonstrated significant improvements in total ASSET score (28.8 vs 16.8; P < .001), checklist score (12.6 vs 7.1; P < .001), and ASSET Safety score (3.9 vs 2.6; P < .001). These results demonstrate that low-fidelity ankle arthroscopy simulation training can improve basic surgical skills, efficiency of movement, and anatomic recognition. The results suggest greater patient safety during ankle arthroscopy following simulation training. Level I, prospective comparative study. © The Author(s) 2015.