Some innocent suspects rely on the memory of strangers to corroborate their alibis. However, no research has examined whether such potential alibi corroborators can accurately recognize an innocent suspect with whom they previously interacted. We developed a novel alibi corroboration paradigm in which undergraduate students (representing innocent suspects who would later provide an alibi) interacted with naïve university employees (representing potential alibi corroborators). Each student briefly interacted with a different naïve university employee (n = 60), and were also each yoked to a different employee with whom they did not interact (n = 60). Employees were presented 24 hours later with either a single photograph of the student or a six‐person array containing a photograph of the student and were asked if they recognized anyone. The majority of employees failed to make a correct recognition of the student. False recognitions, however, were rare. Students exhibited overconfidence that they would be recognized. Findings imply that innocent suspects who rely on strangers to corroborate their alibis may be at risk. Copyright © 2016 John Wiley & Sons, Ltd.