The present study investigated the precise nature of crashing memory reports: Are they truly memories or are they based on beliefs? We asked 88 individuals whether they had seen non-existent footage of the Pim Fortuyn assassination and conducted thorough post-experimental interviews. Two-thirds of our participants falsely reported having seen the footage, while less than 10% also reported details that they could not have seen. Moreover, plausibility ratings of having seen the images were higher than false belief ratings, which in turn were higher than false memory ratings. After having been fully debriefed, 81% of the participants who reported crashing memories attributed their false report to their lack of a full understanding of the critical question. Another 10% of this subsample stated that they truly remembered the images (i.e. false memories). Thus, only a small subset of crashing memory reports seems to be induced by false beliefs and/or false memories. Copyright © 2009 John Wiley & Sons, Ltd.