BACKGROUND: As we move toward competency-based education in medicine, we have lagged in developing competency-based evaluation methods. In the era of minimally invasive surgery, there is a need for a reliable and valid tool dedicated to measure competence in video-assisted thoracoscopic surgery. The purpose of this study is to create such an assessment tool, and to explore its reliability and validity. METHODS: An expert group of physicians created an assessment tool consisting of 10 items rated on a five-point rating scale. The following factors were included: economy and confidence of movement, respect for tissue, precision of operative technique, creation and placement of ports, localization of pathologic tissue, use of staplers, retrieval of tissue in bag and placement of chest tube. Fifty consecutive thoracoscopic wedge resections were recorded and assessed blindly and independently by two experts using the tool. RESULTS: Four residents, four fellows and five consultants performed 1-10 (median 4) operations each. The fellows performed significantly better than the residents (P = 0.03; effect size, ES = 0.72). The consultants scored 11% higher than the fellows, but this difference was not significant (P = 0.10, ES = 0.64). The inter-rater reliability was acceptable (Cronbach's alpha 0.71). CONCLUSIONS: This tool for assessing performance in thoracoscopy is reliable and valid. It can provide unbiased feedback to trainees, and can be used to evaluate new teaching curricula, i.e. simulation-based training. Furthermore, it has potential to aid in certification of new thoracic surgeons.