<p>Prior research demonstrates that eyewitness memory is susceptible to misinformation. Specifically, memory for an original event can be contaminated by post-event information. Recently, we found that susceptibility to misinformation is reduced when mock eyewitnesses are given a warning about the threat of misinformation either before exposure to the post-event information (pre-warning) or after exposure to the post-event information (post-warning). In the present study, we investigated whether the timing of the warning (pre-warning vs. post-warning) and warning frequency (one warning vs. two warnings) impact memory accuracy as well as metacognitive assessments of memory accuracy. In Experiment 1, we found pre- and post-warning similarly decreased the negative impact of misinformation on memory and increased the metacognitive assessments memory. In Experiment 2, repeated warnings (two warnings) also decreased the negative impact of misinformation on memory and increased metacognitive assessments of memory related to misinformation. However, these benefits of repeated warning came at the cost of metacognitive assessments of memory for information that had not been contaminated by misleading post-event information. These results suggest that warnings can improve eyewitness memory accuracy and support the relationship between memory and confidence; however, over-warning an eyewitness may result in under-confidence in accurate memory.</p>