Subscales that measured qualitative features of maternal representations were used in the present study; these were chosen because Zeanah et al. (1994) reported differences between mothers of securely and insecurely attached infants on these scales. These included: (a) Richness of Perceptions, the poverty or richness of representations of the infant and the degree to which the mother ‘‘knows’’ the infant; (b) Openness to Change, the flexibility of the mother ’s representation to accommodate new information about the infant (compared to rigidity); (c) Coherence, the overall organization of the mother’s narrative about the infant and her relationship with the infant and the logical flow of her responses, (d) Caregiving Sensitivity, the degree to which the mother recognizes and responds adequately to the infant’s own needs and experiences including a re- spect for the infant as a separate but dependent individual; and (e) Acceptance, the degree to which the mother is open and accepting of responsibilities involved with adequate caretaking. Higher scores in- dicate higher levels of the construct being measured.