Most published research on eyewitness identification accuracy (a) compares only 2 levels of only one factor (e.g., “sequential” and “simultaneous” levels of the “lineup” factor; “presence” and “absence” levels of the “weapon” factor; “good” and “poor” levels of the “lighting” factor), and (b) relies on results from lab experiments using undergraduate students enrolled in psychology courses (as part of the course requirements). Moreover, most police departments cannot distinguish between potentially reliable versus unreliable eyewitnesses; some criteria for guidance on when they should seek to rely on eyewitness evidence would ensure greater accuracy in eyewitness identifications. We will conduct a series of lab-based and field studies that will: consist of more powerful experimental designs that involve more than a single variable; calibrate the results from lab experiments with those from additional, more realistic “field-type” experiments; investigate possible alternative measures of an individual’s accuracy as an eyewitness (e.g., time between exposure to suspects and selection), and develop simple criteria that can help distinguish potentially valuable from erroneous eyewitness evidence. Our experiments have two goals: (1) examine the (possibly joint) effects of factors regularly cited as forensically meaningful in the literature on eyewitness identification performance; (2) examine a simple measure of face recognition ability that may prove effective at predicting eyewitness lineup identification accuracy.