Purpose ' The purpose of this paper is to test a theoretically derived representation of a relationship management capability. The relationship management capability architecture developed from the literature integrated theory on dynamic capabilities, the resource-advantage theory of competition, and prior capability research in innovation and information technology management. Design/methodology/approach ' The second-order constructs of relationship infrastructure, relationship learning and relationship behaviour argues to represent a relationship management capability (RMC) was assigned measures adapted from the literature, and pilot tested with industry consultants. The final questionnaire was sent to senior executives responsible for customer relationship management in manufacturing and business service firms in the UK. The structural model representing the RMC was shown to be robust with a comparative fit index of 0.91. Findings ' Although the low response rate and the subjectiveness of respondents encourage caution in interpreting the research findings, the results suggest that relationship management systems, implemented through collaborative and flexible behaviours, and renewed through adaptive and generative knowledge derived from experience and challenging current relationship management assumptions, are key dimensions of a RMC. Originality/value ' This framework advances and tests a new theoretical perspective of a relationship management capability that incorporates a capacity for renewal. In addition, it provides managers with a tool to evaluate their organisation's relationship management capability at key stakeholder interfaces on attributes that define relationship infrastructure, relationship learning and relationship behaviour, as this capability is renewed over time.