We assessed the impact of teaching methodological aspects of physics on students’ scientistic beliefs and subject interest in physics in a repeated-measurement design with a total of 142 students of upper secondary physics classes. Students gained knowledge of methodological aspects from the pre-test to the post-test and reported reduced scientistic beliefs, both from their own views and from their presumed prototypical physicists’ views. We found no direct impact of teaching on students’ subject interest in physics. As path analysis indicates, this (...) result can be traced back to opposing paths: Lower scientistic beliefs of students attenuate subject interest while lower presumed scientistic beliefs that they hold of physicists foster subject interest. This finding is in accordance with the self-to-prototype matching theory that predicts an impact of the overlap between students’ self-image and their prototypical image on subject interest in physics. (shrink)