Studies have found that metacomprehension accuracy tends to be poor, but there has also been evidence that rereading a text may improve metacomprehension accuracy. One explanation for this rereading effect is that readers can attend more to the metacognitive level during rereading than the initial reading. However, prior experiments used paradigms that make it difficult to determine if improvements in metacomprehension are due to differences in online comprehension monitoring during reading. Backwards eye movements--regressions--are sensitive to difficulties in comprehension processes, implying a relationship with online monitoring. The current study uses eye tracking to examine online monitoring and its effects on metacomprehension judgments and accuracy in order to further understand the rereading effect. Participants either read texts once or reread texts and made judgments of learning while their eye movements were recorded. The rereading effect was not observed when examining the relationship between judgments of learning and test performance, but a similar rereading effect was observed when examining the relationship between regressions and test performance. The findings suggest that the rereading effect is related to usage of metacomprehension cues related to regressions when making judgments.