<p>Searching for missing or wanted persons is a challenging task that requires sustained attention and active scanning for a difficult-to-recognize stimulus (i.e., an unfamiliar face). Given the naturally low prevalence of missing or wanted persons, people may have low expectations of encountering them in their midst. Understanding how their expectations, combined with feedback and experience, influence search performance is critical for improving real-world search efforts. We manipulated prevalence expectations (40% vs. 2%) and trial-level performance feedback (present vs. absent) in a visual search task for unfamiliar target faces. Critically, the target persons never appeared during the task. We examined how performance changed over time. Among participants who did not receive feedback, those with high-prevalence expectations made more false alarms and terminated their searches earlier than those with low-prevalence expectations. In contrast, participants who received feedback were not affected by prevalence expectations. While prevalence expectations had limited impact on search behavior, feedback enhanced participants’ ability to align their expectations with the true prevalence rate more effectively.</p>