A total of 100 subjects (male=81, female=19) were selected randomly from the Victoria Police (Australia). No specific selection criteria were used. Ten digitized photographs of five doctors and five convicted violent criminals were presented to the subjects. Without knowing that the photographs represented the two categories (criminals and doctors), the subjects rated the personalities of the photographic images using semantic differential scales. Using ratings from 1 to 10, the photographs were rated from unintelligent to intelligent, cruel to caring, dangerous to not dangerous, and hostile to friendly. After this initial observation, police subjects were then told that some of the photographic images were of violent criminals and some were of doctors. This was the "suspicion" phase of the study. The police subjects then reassessed the same 10 photographic subjects. Results indicate that suspicion alone does affect personality rating, based only on facial appearance and no other cues. The experience or the age of the police subjects did not contribute to any bias manifested. 15 tables