Metacognitive monitoring processes have been shown to be critical determinants of human learning. Metacognitive monitoring consist of various knowledge estimates that enable learners to engage in self-regulatory processes important for both the acquisition of knowledge and the monitoring of one's knowledge when engaged in assessment. This study investigated the reliability and the predictive validity of one measure of metacognitive monitoring, "absolute accuracy." "Absolute accuracy" pertains to how accurately students judge their knowledge relative to their actual performance on a test. Students show poor "absolute accuracy" if they judge their knowledge as correct when in fact it is in error, or if they judge their knowledge in error when in fact it is correct. Contrary to previous studies, this study revealed that "absolute accuracy" indices may be reliably measured in an authentic setting. The study also showed that "absolute accuracy" was significantly correlated with concurrent and final exam performance.