Introduction: Attachment theory is a theory of interpersonal behaviour based on whether people feel anxious in relationships (attachment anxiety), and how comfortable they are seeking support from others (attachment avoidance). Attachment anxiety and avoidance have been found to be associated with psychological treatment outcomes and attachment theory has been proposed as an underlying framework for the service delivery design of mental health services. However, how attachment patterns are associated with illness behaviours is unclear. Aims: The aim of this study was to outline and test the use of the conceptual framework of attachment theory in a healthcare setting.Method: This was a mixed methods study. Study 1 was a systematic review of quantitative studies using narrative synthesis to map the associations between attachment and illness behaviours and critically appraise the evidence for the attachment framework. Study 2 was a qualitative study that tested the attachment framework by exploring whether attachment cognitions underpinned expressed behaviours and preferences for treatment of depression. Interviews with patients with diabetes and/ or coronary heart disease and who received psychological therapy within a randomised controlled trial explored illness behaviours. These were self-disclosure about depression, adherence, help-seeking from family or friends, healthcare utilisation and building rapport. A novel approach to qualitative analysis was employed to examine whether individual accounts reflected high or low levels of attachment anxiety and avoidance. Results: The systematic review found that out of 8 studies which explored associations between attachment patterns and illness behaviours, few had strong methodological quality ratings. Attachment avoidance was associated with adherence and healthcare utilisation, attachment anxiety with healthcare utilisation only. The qualitative study findings suggested that healthcare utilisation, self-disclosure and coping mapped onto the attachment avoidance framework. Adherence behaviours and building rapport with different health professionals did not map well onto the attachment framework. A range of clinical, social and treatment related beliefs were found to underpin illness behaviours. Attachment related cognitions were related to beliefs about others as unresponsive. There was no evidence for associations between attachment patterns and treatment preferences in collaborative care. Conclusions: The results suggested that attachment may be associated with healthcare utilisation. Implications are discussed in terms of policy for personalisation of services. There is a need for prospective quantitative research studies testing the association between attachment and healthcare utilisation and in research exploring patients' reports of treatment preferences for different types of psychological interventions. Research in this area would help us determine whether designing services taking account of attachment theory has any patient reported benefit.