Corrected misinformation often continues to influence reasoning; this is known as the continued-influence effect (CIE). It is unclear whether this effect also occurs in impression formation, with some arguing that person impressions are readily updated. The present study tested if a retracted allegation influences person impressions. Participants received examples of behaviors that a fictitious person had allegedly engaged in. The set did or did not include a domestic-violence behavior, which subsequently was or was not retracted. Discredited misinformation was found to influence neither trait ratings of the person nor behavior predictions. This held even when the person's name implied a cultural background stereotypically associated with domestic violence. This provides evidence that under some circumstances, people can fully discount discredited misinformation when building person impressions. However, there was some tentative evidence that corrected misinformation did influence a more indirect measure, namely ratings of the fictitious person's face.