Background. Mild or subclinical hypothyroidism [raised thyroid-stimulating hormone (TSH) but normal free thyroxine (T4)] affects 5–10% of adults. Symptoms are non-specific and TSH levels are needed for diagnosis. Objectives. We explore the relationship between thyroid function and psychological distress and investigate the usefulness of an expert-designed Thyroid Symptom Questionnaire (TSQ) in identifying hypothyroidism. Methods. DEPTH (DEPression and THyroid) is a cross-sectional study of 325 patients recruited from general practices in Bristol, for whom thyroid function tests were requested by the GP. Subjects completed the TSQ, General Health Questionnaire (GHQ-12) and Patient Health Questionnaire (PHQ) and had blood tests for TSH and free T4. Results. The mean age was 45.7 years; 252 subjects (78%) were female; median TSH was 1.6. Psychological morbidity in this population is high: 54.2% have a GHQ-12 score >3, indicating psychological distress. We found no relationship between TSH and psychological distress [adjusted odds ratio 1.02 (95% confidence interval 0.91–1.13), P = 0.78]. The prevalence of hypothyroidism was 6.2% (95% confidence interval 3.8–9.5%). We found no evidence of an unadjusted association between TSQ score and subclinical hypothyroidism [adjusted odds ratio of 1.09 (95% confidence interval 0.95–1.24), P = 0.23]. Conclusions. Those referred for thyroid function tests, although no more likely than others to have hypothyroidism, have high rates of psychological distress. When mild (subclinical) hypothyroidism is detected in patients with psychological distress, it is important that GPs are aware that this is likely to be coincidental rather than causal and offer appropriate treatment.