Interactive Strategy Training for Active Reading and Thinking (iSTART) is an intelligent tutoring system that supports reading comprehension through self-explanation (SE) training. This study tested how two metacognitive features, presented in a 2 x 2 design, affected students' SE scores during training. The "performance notification" feature notified students when their average SE score dropped below an experimenter-set threshold. The "self-rating" feature asked participants to rate their own SE scores. Analyses of SE scores during training indicated that neither feature increased SE scores and, on the contrary, seemed to decrease SE performance after the first instance. These findings suggest that too many metacognitive prompts can be detrimental, particularly in a system that provides metacognitive strategy training. [This paper was published in: X. Hu, T. Barnes, A. Hershkovitz, & L. Paquette (Eds.), "Proceedings of the 10th International Conference on Educational Data Mining" (pp. 404-405). Wuhan, China: International Educational Data Mining Society.]