The psychometric properties of the Italian version of the Hospital Anxiety and Depression Scale and its utility as a screening instrument for anxiety and depression in a non-psychiatric setting were evaluated. The questionnaire was administered twice to 197 breast cancer patients randomised in a phase III adjuvant clinical trial: before the start of chemotherapy and at the first follow-up visit. The presence of psychiatric disorders was evaluated at the follow-up visit using the Structured Clinical Interview for DSM-III-R in 132 patients. Factor analyses identified two strictly correlated factors. Crohnbach's alpha for the anxiety and depression scales ranged between 0.80 and 0.85. At follow-up, 50 patients (38%) were assigned a current DSM-III-R diagnosis, in most cases adjustment disorders (24%) or major depressive disorder (10%). Receiver operating characteristics (ROC) analysis was used to test the discriminant validity for both anxiety and depressive disorders. The comparison of the areas under the curve (AUC) between the two scales did not show any difference in identifying either anxiety (P = 0.855) or depressive disorders (P = 0.357). The 14-item total scale showed a high internal consistency (alpha = 0.89 and 0.88) and a high discriminating power for all the psychiatric disorders (AUC = 0.89; 95% CI = 0.83-0.94). The cut-off point that maximised sensitivity (84%) and specificity (79%) was 10. These results suggest that the total score is a valid measure of emotional distress, so that the Italian version of HADS can be used as a screening questionnaire for psychiatric disorders. The use of the two subscales as a 'case identifier' or as an outcome measure should be considered with caution.