Background. Students perspective on education is of crucial importance for its effectivity, but students opinions are seldom acknowledged by teachers and designers. Student participation in the educational design process could be a suitable tool to better take students preferences into account. However, for effective participatory design, it is necessary to know whether students have stable preferences for the design of their education. Changeability of preferences would require a more continuing design process allowing continuous adaptations. Aims. This longitudinal survey study aimed to determine the changeability over time of students preferences for different aspects of a learning environment. Additionally, causes of possible changes in preferences are investigated. Sample. The participants were 1,335 high school students of five schools for secondary education in the Netherlands, joining this study during a period of 2 years. Method. Data about students preferences were collected at three moments, using the Inventory of Perceived Study Environment Extended. Learning-related student characteristics, such as processing strategies and motivational orientations, were measured with the Inventory of Learning Styles. Additionally, data on learning performances were collected. Results. The results showed stability on preferences for almost all studied characteristics of the learning environment. Particularly remarkable was a drop in desirability for student autonomy. This was larger for students with a certificate-oriented motivation and smaller for self-regulated students. Additionally, poorly performing students had a larger decrease in preference for autonomy. Conclusions. The stability on most aspects supports that participatory design might result in fairly stable instructional designs, although caution is needed with respect to student autonomy.