This thesis outlines a body of research relating to the concept of need and needs-led health service responses for people with intellectual disabilities (ID). In summary, it describes the content, linkages, strengths and limitations of seven published research papers, each with a different, but related focus/question about need. The use of diagnoses in the field of mental health and ID is limited, leading some healthcare providers to adopt a more needs-based approach. Need, however, can be conceptualised in a variety of ways, each with its own benefits and drawbacks. Arguably the most objective of these are normative needs (i.e. a professional's assessment of an individual's need against a notional standard). This objectivity, combined with the limitations of diagnoses, has led some parts of the NHS to adopt normative needs assessment as the basis of a new payment system. Initially, the existing needs assessment tool was critiqued and recommendations for its improvement published. Outputs from the original tool's use (clusters of people with similar mental health needs) were also empirically confirmed to have logical relationships with the established diagnostic taxonomy. Subsequently the tool was adapted and validated for use in specialist ID healthcare settings. Analyses of data resulting from this new tool's use identified statistically robust groups/clusters of individuals with similar severities and combinations of objective needs. These clusters were clinically recognisable and differed according to key features such as severity of ID, challenging behaviours, Autism symptoms and physical health problems. Secondary analysis of these data then confirmed an existing clinical hypothesis i.e. that challenging behaviours in people with more severe ID may be viewed as behavioural manifestations of underlying mental health problems. Finally, these analyses suggested the normative needs assessment tool had utility in identifying users of community-based specialist ID services most at risk of admission to a specialist ID hospital.