The unreliability of eyewitness evidence is, at first glance, counter intuitive - after all, what better proof could there be than a person who has seen the crime firsthand? After reading several journal articles in Psychological Science and Law and Human Behavior, I have found that eyewitness accounts are essentially recollections from memory, and memory is fallible. Psychologist Elizabeth Loftus suggests that our original memories can be supplemented with new information or even modified with inconsistent information. In an experiment, Loftus, Miller, and Burns (1978) had participants view slides of a car accident at an intersection with a stop sign. Immediately after viewing these slides, the participants answered a set of questions. Half of the participants read a question that included the consistent phrase “stop sign,” whereas the other half read an inconsistent question with the phrase “yield sign.” Then, the participants were shown pairs of slides and asked to choose which in the pair they had seen earlier. Participants who received the inconsistent information about the yield sign in the questions were more likely to have said that they saw the slide with the yield sign, compared to those who did not receive the inconsistent information. These findings imply that memory may not be the most reliable tool for conviction, and hence eyewitness identification procedures should be revised and carefully carried out.