Background Postnatal maternal depressive symptoms are consistently associated with reduced quality of mother–infant interaction. However, there is little research examining the role of maternal cognitive factors (e.g. rumination) in the relationship between depressive symptoms and mother–infant interaction quality. This study investigated the hypotheses that: dysphoric mothers would demonstrate less sensitive behaviour towards their infants compared with nondysphoric mothers; mothers induced to ruminate would be less sensitive towards infants; rumination would moderate the relationship between maternal depressive symptoms and maternal sensitivity and the impact of the rumination induction would increase following a stressor (still face) task. Method Mothers (N = 79; 39 dysphoric and 40 nondysphoric) and their infants were randomised to either a rumination induction or a control condition. Maternal sensitivity in mother–infant interactions was assessed before and after the induction using the CARE Index. In the second interaction task, mothers also completed the still‐face procedure as a stressor. Results Extending previous research, mixed measures ANOVAs demonstrated that dysphoric mothers had reduced quality of interaction with their infant compared with nondysphoric mothers and that mothers in the rumination condition exhibited reduced sensitivity towards their infants relative to mothers in the control condition. Further, maternal sensitivity worsened further after the still‐face procedure in the rumination condition, but not in the control condition. Conclusions This study suggests that the repetitive, internal focus of a ruminative state is causally implicated in mother–infant interaction quality, regardless of the level of depressive symptoms. This research extends understanding of specific mechanisms involved in the quality of the mother–infant relationship.