The General Health Questionnaire has had some popularity as an index of minor psychiatric morbidity and was used in the present study to ascertain the emotional state of newcomers to a practice in a new town. High scorers on the GHQ had more episodes of illness, had more severe ratings of psychological problems, and were more likely to receive a formal psychiatric diagnosis than were low scorers. A second survey one year later confirmed the variability of response to the GHQ inherent in a 'present state' inventory. Doubts are expressed as to the psychiatric nature of the emotional upset measured by the GHQ.