This study examined predictors of stability and change in women's maternal representations of their children. Participants were 180 women, recruited from the community, half of whom had experienced domestic violence during pregnancy. Maternal representations of were assessed with the Working Model of the Child Interview (WMCI; Zeanah, Benoit, Hirshberg, Barton, & Regan, 1994 Zeanah, C. H. , Benoit, D. , Hirshberg, L. , Barton, M. L. and Regan, C. 1994. Mothers' representations of their infants are concordant with infant attachment classifications. Developmental Issues in Psychiatry and Psychology, 1: 9–18. [Google Scholar] ) during the last trimester of pregnancy and again at the child's first birthday. Results indicated that when collapsed into balanced and non-balanced categories, 71% of the sample was stable over time, and women who had balanced representations had significantly more stable representations than women who had non-balanced representations (p < .001). Income, single parenthood, abuse status, and depressive symptomatology predicted change. In addition, women who became non-balanced postnatally benefited from having balanced representations while pregnant to buffer the quality of their interactions with their children.