Fostering student motivation is an important educational goal. However, motivation in the classrooms is rather heterogeneous, particularly in mathematics and physics. This study examines the potential of (textbook) tasks to promote student motivation. Based on self-determination theory (SDT) and theory of interest, a low-inference coding scheme was developed and validated by applying the framework of item response theory (IRT) to assess the motivational potential of tasks. Current ninth grade mathematics and physics tasks (N = 254 task units) were analyzed using the categories differentiated instruction, real-life context, autonomy support, competence support, and support for relatedness. Additionally, differences between mathematics and physics tasks were examined. Results indicate the coding scheme’s high interrater reliabilities and empirical validity. Furthermore, we found only a low occurrence of motivational features in mathematics and physics tasks, with few subject-specific differences in favor of mathematics. The coding scheme can contribute to optimizing motivation-supportive instructional designs.