This study investigated effects of misleading post‐event information, delay, and centrality definition on eyewitness memory and suggestibility for a negative event (a vividly filmed murder). Either immediately or 2 weeks after viewing the film, 93 adults read a (misleading or control) narrative about the event and then completed a recognition memory test. Misinformation acceptance was operative, but strong evidence for memory malleability was lacking. Compliance predicted misinformation effects, especially on the delayed test. Although accuracy was generally higher for central than peripheral information, centrality criteria influenced the pattern of results. Self‐report of greater distress was associated with better recognition accuracy. The results suggest that use of different centrality definitions may partly explain inconsistencies across studies of memory and suggestibility for central and peripheral information. Moreover, social factors appeared, at least in part, to influence misinformation effects for the highly negative event, especially as memory faded. Implications for eyewitness memory and suggestibility are discussed. Copyright © 2013 John Wiley & Sons, Ltd.