PURPOSE: To investigate the impact of a proficiency-based progression (PBP) curriculum employed to teach trainees in the skills needed to demonstrate proficiency for an arthroscopic Bankart repair (ABR) and an arthroscopic rotator cuff repair (ARCR) by objectively comparing pre- and immediate post-course performances. METHODS: In a prospective study, 16 arthroscopy / sports medicine fellows and 2 senior residents (Complete Group: N=18) were randomly assigned to perform a pre-course cadaveric ABR (Bankart Subgroup: N=6), ARCR (Cuff Subgroup: N=6), or basic skills on a shoulder simulator (N=6). After completing a PBP training curriculum, all 18 registrants performed both an ABR and ARCR scored in real time by trained raters using previously validated metrics. RESULTS: The Bankart Subgroup made 58% fewer objectively assessed errors at the completion of the course than at baseline (P = 0.004, CI -1.449 to -0.281) and performance variability was substantially reduced (SD = 5.89 vs. 2.81). The Cuff Subgroup also made 58% fewer errors (P = 0.001, CI -1.376 to 0.382) and showed a similar reduction in performance variability (SD = 5.42 vs. 2.1). Only one subject's pre-course baseline performance met the proficiency benchmark compared with 89% and 83% of the all registrants on the final ABR and ARCR cadaveric assessments respectively. CONCLUSIONS: The results of this study reject the null hypothesis. They demonstrate that the implementation of a PBP simulation curriculum to train the skills necessary to perform arthroscopic Bankart and rotator cuff repairs results in a large and statistically significant improvement in the trainee's ability to meet the two related performance benchmarks. Proficiency was demonstrated by 89% and 83% of the trainees for an ABR and an ARCR respectively, in a two- and one-half day course.