This study examines how often depression and anxiety, in patients with diabetes, are detected by health professionals; and whether detection is influenced by patient characteristics (age, gender), illness factors (duration of illness, diabetes control), and self-reported levels of depression and anxiety. Prevalence rates of clinically significant depression and anxiety were high (57% and 36%, respectively); however, of those identified, only 44 and 36 per cent, respectively, were detected by staff as depressed or anxious. The only significant predictors of detection were severity of depressive and anxious symptoms. Patient and illness characteristics did not influence whether professionals identified emotional problems in their patients.