The main goal of the present research was to test whether the presence of analogies would affect the relative accuracy of metacognitive judgments about learning from expository science texts, and whether any effect would depend on the type of cues that readers used as the basis for their judgments of comprehension. In a series of experiments, students read texts that either contained or did not contain analogies; were asked to judge how well they understood each text; took comprehension tests for each topic; and were asked to self-report the basis for their judgments. Relative metacomprehension accuracy was computed as the intra-individual correlation between judgments and test performance. Results showed that the presence of analogies can lead to poor relative metacomprehension accuracy for students who fail to use situation-model-based cues to judge their understanding of text. [This paper was published in "Learning & Instruction" v55 p113-123 2018.]