It is much easier to look for an unfamiliar character (upsidedown A) embedded among familiar ones than to look for a familiar character (A) among unfamiliar ones. Furthermore, the nature of the background seems more important to performance than the nature of the target. The basic experiments involved showing subjects a matrix with nine characters. On target-present trials, eight of the characters were background items and one character was a target item. On target-absent trials, all nine matrix positions were filled by background characters. The types of unfamiliar characters used were rotated English letters, partial letters, and Gibson figures. The familiar characters were upright English letters or digits. Search was easier through familiar backgrounds than through unfamiliar backgrounds with all of the character types used and whether measuring speed or accuracy.