Stereotypes of battered women invoked by photographs and videotapes were examined. 'Informed' participants were told some women they would judge were battered women, and were asked to select from a pool of 10 the one they believed was most or least likely to be a battered woman. Participants rated their choice's attractiveness, likeability, and distinctiveness. Finally, the informed participants sorted 10 additional women into the categories 'battered' and 'non-battered.' A separate group of 40 'uninformed' participants rated the same 20 women in terms of the three traits, without knowing that some women were actually battered. Results showed that certain women were consistently chosen as category exemplars, and that participants were poor in their categorizations. Ratings for women judged likely to be battered were less positive than those for women judged unlikely to be battered. Women judged likely to be battered were rated less positively by informed participants than by uninformed participants. There were no consistent differences between photographs and videotapes.