Observing maintenance competencies in training settings is difficult. Many critical attitudes such as safety-awareness or team skills are covert and occur irregularly. As a result, assessing such competences for learning and testing often is subjective and not transparent. TNO and the Royal Netherlands Air Force (RNLAF) together have created a method for learning and assessing the development of such critical attitudes and skills, called the Visual Assessment Method for Maintenance training (VAMM). In this method, video-cameras are set up in the workshop, monitoring a team of learners doing realistic F16 maintenance tasks. One of the team members is responsible for evaluation of task performance. A commercial computer-based tagging tool (Darthfish) is used to mark incoming video events, and quickly classify them according to predefined evaluation criteria such as `bookwork discipline' or `team co-operation'. A selection of marked video-episodes is used in an After Action Review (AAR), showing examples of good and bad displays of attitudes and skills. Both task performers and evaluator can learn from this reflection process. Instructors may apply the method in exam settings, providing an objective and transparent manner of assessing realistic tasks. The VAMM method is currently being tested by means of a pilot project at a RNLAF school with two teams of four learners. The first results show that learners are capable to assess mutual task performance by means of video analyses; (a) a sufficient number of events was tagged for proper AAR and examinations and (b) the tagged events covered the most important attitudes and skills relevant for maintenance training to be learned. It can be concluded that the VAMM method has great potential in the maintenance training domain. Currently, an experimental design study is set up to test the impact of the method more rigorously.