Pezdek et al. (2006) reported that although imagining a plausible event increased people's belief in the event, imagining an implausible event did not. In response, Rubin and Berntsen (2007) conducted a survey and reported that only 17.8% considered it implausible that someone ‘with longstanding emotional problems and a need for psychotherapy’ could be a victim of childhood sexual abuse and forget the abuse. We replicated but qualified their findings; perceptions of the plausibility of this event for (a) respondents themselves and (b) other people in their cohort were substantially lower than the perceived general plausibility reported by Rubin and Berntsen. These findings limit the generalizability of Rubin and Berntsen's results to perceptions of personal plausibility and cohort plausibility, even for individuals indicating that they are likely to seek psychotherapy. Consequently, the risk of inducing false memories in psychotherapy may not be a ‘substantial danger’ as Rubin and Berntsen suggest. Copyright © 2008 John Wiley & Sons, Ltd.