Currently, a contradiction exists between theories of motivation, which emphasize the positive influence of academic challenge on student motivation, and research, which has found contradictory results regarding this relationship. The purpose of this dissertation was to investigate the perceived and actual effects of academic challenge on students' motivation through a multi-perspective and multi-methodological approach. Study I examined middle school teachers' views about implementing challenging instruction while participating in a whole-school professional development initiative. Current education policy and reform advocate for increasing the level of challenge in K-12 classrooms in order to maximize students' learning and academic success. A grounded theory analysis revealed teachers' feelings about challenge, the pressures that affected their decision to implement challenging instruction, and teachers' use of instructional practices to challenge students. Classroom observations were also analyzed to explore whether teachers' comments were related to differences in their use of challenging instruction. Teachers perceived 19 different pressures related to implementing challenging instruction, with pressures from students the most common across all subject areas. Some teachers were able to resolve pressures from students by having conversations with students about challenge, providing emotional and motivational support, scaffolding students' thinking, and increasing student autonomy. Implications for teachers' practice and professional development are discussed. Study II investigated changes in middle school students' situational interest and affect during a moderately difficult reading task. The aim was to explore how changes in interest (topic and situational) and affect were related to students' fluency throughout the task and perceived difficulty. Interest and affect were recorded at four time points: before reading, twice during reading, and after reading. Latent growth curve analysis showed that interest and affect had different patterns of decline during the task. The change in interest was predicted by perceived difficulty and fluency, whereas the change in affect was predicted only by perceived difficulty. Results of an autoregressive, cross-lagged path model indicated that fluency significantly predicted subsequent ratings of situational interest, and topic interest predicted fluency on the first section. These findings suggest that, in the context of moderate reading difficulty, perceived difficulty and fluency have divergent effects on different motivational outcomes.