Examines audience-tuning effects on memory. Specifically investigates (1) if rather global perceptions and interpretations, such as a co-witness' liking for a suspect, can impact eyewitness memory for a forensically relevant incident and (2) if this impact depends on features of the audience. Participants were 64 university students (mean age 24 years). After observing a video in which the target persons could be perceived and described in varying ways, student communicators tuned their description of an employee to either an equal-status audience (a student temp) or a higher-status audience (a company board member). Though audience-tuning occurred under both conditions, a memory bias was confirmed in the equal-status condition only: Information that was not directly related to the incident had a strong effect on participants' representation of the event. When discovering that their audience favored the target person (also finding him to be less responsible for the incident), participants actually assigned a weaker penalty to him. A simple likability judgment rendered by participants' audience was enough to elicit audience-congruent biases on eyewitness' memory and judgment. These results demonstrate how audience-congruent biases influence eyewitness' memories and judgment.