[Methods: scales, biological and clinical criteria]

Assessing depressive symptomatology in antidepressant drug trials is limited by the absence of external validators of depression. Moreover, a high level of standardisation is required by regulatory authorities for registering new drugs. Consequently, the methods used are relatively old, and recent improvements have been minor. Methods used to assess new antidepressants appear highly standardised and lack objectivity. Improvements to these methods may emerge from the discovery of external validators of depression on the one hand and of drugs with new mechanisms of action on the other.