Despite the high prevalence of shoulder-neck pain in the community, and the fact that it is commonly a persistent and disabling condition, only a minority of sufferers seek medical help. We investigated the association between primary care consultation and subsequent outcome in a cohort of shoulder-neck pain sufferers. A population with unilateral shoulder-neck pain was identified by a questionnaire mailed to 4002 adults randomly selected from the register of one family practice. Subjects were asked to shade areas of pain on a blank manikin, and give demographic details and scales of pain, anxiety, and depression. For the following 2 years, general practitioner (GP) consultations for shoulder and neck problems were determined using the practice database. The persistence of pain and degree of shoulder-specific disability, as well as general health status using the Medical Outcome Study Short Form-36 (SF-36), were assessed by means of a second postal survey at 2 years' followup. Three hundred four subjects (11.7% of questionnaire responders) had unilateral shoulder-neck pain at baseline, and 224 were included in the study analyses. Of these, 47 (21%) consulted their GP for shoulder-neck problems over the 2 years. Of the 47 consulters, 36 (77%) reported shoulder-neck pain at followup; this was a higher percentage than that for nonconsulters (RRadjusted = 1.3). Among all subjects with persistent shoulder-neck pain, consulters were more likely than nonconsulters to have shoulder related disability at followup (RRadjusted = 1.6). On average, consulters had more pain and lower levels of physical functioning at followup than nonconsulters as measured by the SF-36. The minority of shoulder-neck pain sufferers who consult a primary care practitioner do not have better subsequent pain and disability outcomes than those who do not consult. Our findings raise questions about the current influence of medical care on the natural history of this condition.