This study reports the efficacy of the General Health Questionnaire (G.H.Q.) in the secondary prevention of minor psychiatric illness in a primary-care setting. 1093 consecutive attenders at a general practitioner's surgery were screened for minor psychiatric disorder using the G.H.Q. 32% were found to have a conspicuous psychiatric disorder and a further 11% had a hidden psychiatric disorder. The group with hidden disorders were randomly assigned to a treated index group and an untreated control group. The effects of case detection and treatment were beneficial and immediate, with the duration of episode of the disorder being much shorter for patients whose disorder was recognised by the general practitioner. For patients with more severe disorders there are significant differences still demonstrable between the groups one year later; but patients with mild disorders do equally well, some recovering spontaneously but others becoming manifest to the general practitioner over the next year and so receiving treatment. The "detected" group of patients increased their consultations for emotional complaints over the next year, but their total consultation-rate was not increased.