In March 2003, "Psychological Science" published an article by Goodman and colleagues examining memory for prior childhood abuse in a sample of adolescents and young adults who as children had been the victims of childhood abuse that led to criminal prosecution. At the time of the prosecution cases, the children in this study had been part of a research project conducted by Goodman and colleagues (Goodman et al., 1992). This sample has particularly well-documented abuse histories because of the criminal investigation and prior research. For the follow-up research reported by Goodman et al. (2003) in "Psychological Science," participants had been asked to report any experience of childhood sexual abuse. Failure to report the documented abuse could arise because of forgetting. Goodman et al. (2003) reported that 81% of the participants reported the documented abuse, a proportion that is somewhat higher than in some previously published prospective studies assessing memory for abuse (e.g., Williams, 1994). Near the end of the paper, the authors summarize their findings: "Results from this study indicate that forgetting of CSA may not be a common experience, at least not in a prosecution sample." The Goodman et al. (2003) paper is a very important one in a number of regards, including the unique sample that was assessed. Because the participants had been involved as children in criminal cases related to their victimization (a so-called "prosecution sample"), documentation for the abuse was particularly thorough and offered researchers the opportunity to compare later disclosures with documented records. However, along with the fact that the paper provides important new information, there is a substantial risk that the research reported in the paper might be misinterpreted. In particular, the results of this research might be generalized to non-prosecution samples. This inappropriate generalization could dramatically change the interpretation and application of the results reported in the paper. Is forgetting "uncommon" in general among abuse survivors, or rather, is it remarkable that there was such a high rate of non-disclosure in this unusually legitimated and rehearsed sample? The first interpretation seems the likely one to be made from this paper by a superficial reading, in part because nowhere in the title or abstract is it mentioned that the sample is a prosecution sample. This commentary discusses the biases in Goodman and colleagues' sample.