The major source of information in a trial’s search for factual truth is the oral testimony of witnesses. The main worry about those witnesses is that they will be in error or will intentionally lie or mislead. Consequently, the law provides tools to help assess the veracity and accuracy of witnesses. However, the effects of some of those tools (e.g., oath) are unknown, and we know that people in general are bad at using certain other tools (e.g., demeanor cues in assessing veracity). Cross-examination often is praised as the law’s greatest method for getting at both veracity and accuracy, but surprisingly little psychology research exists on it. One worry about cross-examination is the same worry that psychologists have about leading questions: source confusion (jurors sometimes don’t remember which witness said what, and so don’t remember which testimony was discredited during cross-examination). Like the problem of disregarding inadmissible evidence, it can be difficult or impossible to erase what has already been learned. Historically, evidence law precluded people from testifying when the law assumed they would be likely to lie or make mistakes. The rules then opened the courtroom doors wider, allowing nearly everyone to testify, with the assumption that lawyers could expose or jurors could recognize when witnesses were lying. Interestingly, in the realm of privileges, which prevent some potential witnesses from testifying for various psychological and policy reasons, end up precluding many of the potential witnesses who would have the most incentive to lie on the witness stand.