Although a large body of evidence points to an association between postnatal depression and non-optimal mother-infant interactions, empirical research findings are mixed. This study investigated the impact of maternal postnatal depression, depression chronicity, and infant gender on mothers' behaviour and speech in interaction with their 15-month-old infants in a sample of generally well-educated, middle-class mothers. When mothers who had experienced postnatal depression (n=77) were compared with mothers who had not (n=35), few differences in the quality of interactions were found. However, mothers who had experienced postnatal depression were more likely than never depressed mothers to demonstrate signs of intrusiveness in interaction with their infants. Neither depression chronicity nor infant gender modified this association. Results are discussed with respect to the characteristics of the sample, measurement issues and models of cumulative risk.