Although 'patient-centred' consulting skills are increasingly seen as crucial for the delivery of effective primary care, there is significant lack of clarity over the precise definition of the term, optimal methods of measurement, and the relationship between patient-centred care and patient outcomes. The present study sought to review all empirical studies to date that have investigated the relationship between measures of patient-centred consulting and outcomes in primary care, and to examine the methodological rigour of the studies. A number of observational studies were identified, all of which reported some relationships between doctor behaviour defined as 'patient-centred' and a variety of patient health outcomes. However, the pattern of associations was not clear or consistent, and some of the studies had shortcomings in terms of their internal and external validity. Although the current evidence base may be suggestive of a relationship between patient-centred consulting behaviour and patient outcomes, the case has not been made definitively.