The authors document the pervasiveness of this problem and compare the relative strengths and weaknesses of more versus less realistic research methods. Several recurrent themes of methodological criticism are identified and discussed. The central conclusion is that a number of methods and settings, including highly 'unrealistic' ones (e.g., laboratory and simulation studies), can inform researchers about courtroom behavior. The authors reject the growing counsel to limit the range of 'acceptable' methods for studying courtroom psychology. Findings derived from unrealistic simulations should be appropriately qualified and should not be forwarded as the primary basis for policy changes. Over 100 references, footnotes, and tables are included.