Metacomprehension monitoring accuracy is defined as the ability to accurately predict how well one will do on a later test of learned material. Metacomprehension monitoring is presumed to be a critical skill for the effective self-regulation of study behaviors that impact learning. In two experiments, ecologically valid science texts and inference tests were employed to examine whether a test expectancy intervention could improve students’ metacognitive judgments, self-regulated study, and learning outcomes. Experiment 1 was a lab experiment in which test expectancies were instilled only after reading was complete, thus preventing any encoding effects. Results suggest that test expectancies impact metacomprehension monitoring accuracy via selection of more valid cues at the time of judgment rather than only via encoding effects that impact cue accessibility. Experiment 2 was a classroom study showing that the effect of test expectancies on monitoring accuracy translates into more effective self-regulated study and improved learning.