A self-administered questionary (the General Health Questionnaire) aimed at detecting current psychiatric disturbance was given to 553 consecutive attenders to a general practitioner's surgery. A sample of 200 of these patients was given an independent assessment of their mental state by a psychiatrist using a standardized psychiatric interview. Over 90% of the patients were correctly classified as "well" or "ill" by the questionary, and the correlation between questionary score and the clinical assessment of severity of disturbance was found to be +0.80.The "conspicuous psychiatric morbidity" of a suburban general practice assessed by a general practitioner who was himself a psychiatrist and validated against independent psychiatric assessment was found to be 20%. "Hidden psychiatric morbidity" was found to account for one-third of all disturbed patients. These patients were similar to patients with "conspicuous illnesses" in terms both of degree of disturbance and the course of their illnesses at six-month follow-up, but were distinguished by their attitude to their illness and by usually presenting a physical symptom to the general practitioner.When 87 patients who had been assessed as psychiatric cases at the index consultation were called back for follow-up six months later, two-thirds of them were functioning in the normal range. Frequency of attendance at the surgery in the six months following index consultation was found to have only a modest relationship to severity of psychiatric disturbance.It is argued that minor affective illnesses and physical complaints often accompany each other and usually have a good prognosis.