This longitudinal study aimed to investigate the impact of chronic and transient maternal depression on children's attachment representations at 4 years of age measured with the Attachment Story Completion Task (Bretherton, Ridgeway, & Cassidy, 1990). The impact of concurrent maternal depressive symptoms was also considered. A secondary aim was to investigate the continuity of attachment classification at 15 months with child attachment representations at 4 years. Children of mothers who were concurrently depressed were more likely to have attachment representations characterized by uncontained physical aggression. In contrast, there were no significant relationships between exposure to prior chronic maternal depression and children's attachment representations at 4 years, nor were there any significant relationships between behavioural assessments of attachment during infancy and later representational assessments. Findings are discussed in the context of caregiving behaviours that may foster secure attachment relationships and influence a child's social-emotional development, limitations of representational measures, as well as poor attachment stability.