In the typical crime, victim or bystander eyewitnesses are interviewed several times under varying conditions. Usually the first time is shortly after the crime by a uniformed police officer; a second time can be an hour or two later by a more experienced detective likely at a more leisurely pace and at a different location. A third time may be several days or weeks after the incident usually by a prosecuting attorney in his or her office. Depending on the case, additional rounds of pre- trial or trial interviews may occur several months to years later. In all, the typical witness is interviewed about a half dozen times by different people, under different conditions, and for radically different purposes. Despite variability in these interviews, inconsistency in recollections is very effective in discrediting a witness. Yet, we know little about what such inconsistencies indicate about the accuracy of a witness's testimony. Should judges or juries give greater credibility to witnesses who are consistent? How relevant is consistency of testimony to the accuracy of recollection? Drs. Fisher and Cutler plan to examine the relationship between eyewitness consistency of recollection and accuracy through undertaking three laboratory experiments. Their model postulates that the relationship between the two can be either strongly and positively correlated, weakly correlated, or even negatively correlated, and the experiments derived from this model test this relationship empirically. Each of the experiments manipulates legal relevant and model-driven variables (e.g., encoding quality, similarity of recall conditions, peripheral versus central information). For purposes of generalizability, in the first two experiments, a criminal event is witnessed; whereas in the third experiment a non-criminal event is observed. These investigators hypothesize that the relationship between accuracy and consistency will vary. This counterintuitive finding would be of considerable value to an understanding of eyewitness behavior and testimony. More generally, the work should advance knowledge about (1) the effects of retrieval operations on recollection, and (2) the significance of consistency of recall.