BACKGROUND: Balancing the normative expectations of others (accountabilities) against the personal and distributed resources available to meet them (capacity) is a ubiquitous feature of social relations in many settings. This is an important problem in the management of long-term conditions, because of widespread problems of non-adherence to treatment regimens. Using long-term conditions as an example, we set out middle range theory of this balancing work. METHODS: A middle-range theory was constructed four stages. First, a qualitative elicitation study of men with heart failure was used to develop general propositions about patient and care giver experience, and about the ways that the organisation and delivery of care affected this. Second, these propositions were developed and confirmed through a systematic review of qualitative research literature. Third, theoretical propositions and constructs were built, refined and presented as a logic model associated with two main theoretical propositions. Finally, a construct validation exercise was undertaken, in which construct definitions informed reanalysis of a set of systematic reviews of studies of patient and caregiver experiences of heart failure that had been included in an earlier meta-review. RESULTS: Cognitive Authority Theory identifies, characterises and explains negotiation processes in in which people manage their relations with the expectations of normative systems - like those encountered in the management of long-term conditions. Here, their cognitive authority is the product of an assessment of competence, trustworthiness and credibility made about a person by other participants in a healthcare process; and their experienced control is a function of the degree to which they successfully manage the external process-specific limiting factors that make it difficult to otherwise perform in their role. CONCLUSION: Cognitive Authority Theory assists in explaining how participants in complex social processes manage important relational aspects of inequalities in power and expertise. It can play an important part in understanding the dynamics of participation in healthcare processes. It suggests ways in which these burdens may lead to relationally induced non-adherence to treatment regimens and self-care programmes, and points to targets where intervention may reduce these adverse outcomes.