Most research on battered women concerns women of working class and middle class backgrounds. This study examined differences between poor and non-poor battered women and ways in which poverty mediates the experience of intimate violence. Subject data were obtained on 57 cases of battered women who killed their abusers between 1978 and 1984. Statistical tests were conducted to compare the 30 poor subjects to the 27 non-poor subjects. The results revealed that, compared to non-poor subjects, poor subjects were more likely to be non-white, undereducated, have violence in their family backgrounds, report that they killed their abuser while an attack was ongoing, and have used knives to kill their abuser. They were less likely to have been legally married to their abuser, to have used a gun to kill their abuser, and to have had expert testimony at their trials. These findings illustrate the existence of both demographic and experiential differences between the poor and the non-poor battered women who killed their abusers. In spite of the unlikely inclusion of expert testimony at trials of poor women, rates of acquittal at trial were not statistically different, possibly due to the finding that poor women used violence at a time that was easier for the jurors to understand as self-defense. (NB)