The association between disruptive behaviours in children and depression in mothers is well-established (Cummings & Davies, 1994). Both are associated with social isolation, socio-economic deprivation, marital conflict, and overgeneral recall of memories. Maternal depression is also associated with poorer treatment outcome in parent-training programmes (Webster-Stratton & Herbert, 1995), although some parent-training intervention studies have reported improvements in maternal depression following intervention (e. g. Hutchings, 1996a, 1996b). Given some similarities between parent-training programmes and cognitive-behaviour therapy (CBT) techniques, this pilot study proposed to investigate the relationship between a CBT intervention for depressed mothers and its effects on the maternal reporting of disruptive behaviours in their children in comparison to an observational measure. The combination of measures used were Beck Depression Inventory-2°d Edition, Beck Anxiety Inventory, General Health Questionnaire-30, Eyberg Child Behaviour Inventory, Child Behaviour Checklist, Parenting Stress Index, Autobiographical Memory Test, Community Contacts Questionnaire, and The Brown Circles Task with Standardised Observation Codes -III B as the observational measure. Ten depressed mothers referred to an Adult Mental Health team for treatment, who also had children between the ages of 2 and 7 years who were reported to be exhibiting clinically significant disruptive behaviours (as measured by ECBI) participated in the study. Demographic data suggested that the sample were a `high risk' group for the development of depression and disruptive behaviours. Following the CBT intervention both self-reported depressive symptoms and maternal report of child disruptive behaviours decreased significantly on all standardised measures. However, no significant differences were shown on the observational measure in either maternal or child behaviours and interactions, although child behaviours changed in the predicted direction. Supportive social networks were predictive of better outcome for maternal depression (as measured by BDI-II). The limitations of the study, and the clinical implications of the results are discussed.