<p>Numerous studies indicate that students’ metacognitive monitoring of their learning can be biased by factors related to the study experience, but less is known about how performance on an unrelated task affects judgments of learning (JOLs) on a subsequent learning task. In four experiments, college students completed either a very difficult or very easy math task (or no math task). All participants then completed the same paired-associates learning task, making a JOL as they studied each item. In all four experiments, participants judged their learning of the paired-associates to be better after completing the easy math task than after completing the difficult math task. This occurred even though participants’ performance on the memory task was not affected by which math task they completed, and occurred with and without feedback on the math task. Differences in JOLs led to differences in restudy time. Difficult math lowered state self-esteem, but it did not mediate the effect on JOLs. The results indicate that performance on a prior, unrelated task can bias students’ JOLs on a subsequent learning task. Such effects are not explained by current theories of metacognitive monitoring, but suggest that students who are studying for multiple exams close together in time (e.g., final exams) might inappropriately judge their learning for one topic after experiencing a test in another topic.</p>