The way in which people view reasons why women stay in abusive relationships has important implications for prevention and treatment of abused women as well as for community education. Numerous investigations have focused directly on the elements influencing a battered woman’s decision to stay, however, little research has focused on observers’ views of the specific reasons why women remain in abusive relationships. It was hypothesized that sympathy and justifiability would be related constructs in the degree of sympathy due the battered woman for the reason(s) she remained. Reasons that were perceived to be externally imposed on the battered woman were expected to be viewed as deserving more sympathy and as more justifiable for the woman remaining for those reasons. Two hundred thirty-three undergraduates at a southeastern university were recruited as participants in the present study. Results indicated that the reasons women remained in abusive relationships included dimensions of sympathy and justifiability, but that very few of the reasons were consistently viewed as very sympathetic or highly justified. Although some reasons were viewed in a more sympathetic light and were seen as providing a reasonable explanation as to why the woman was still in the abusive relationship, the participants as a group were not likely to completely excuse or have very strong sympathy for the battered woman, even for the direst reasons. Physical restraint, physical threat, and serious lack of help/resources were viewed by observers as being the most sympathetic and the most justifiable for the woman remaining in the relationship. These results implied that more sympathy would be felt for the woman and more justification would be afforded to her reasons for staying if she were viewed as having little control or ability to change the situation in which she found herself. Public education regarding battered women’s dilemmas and experiences could promote greater awareness for those who treat or come into contact with battered women. 4 tables, 44 references