Paper and pencil screening tests for depression have never become widely used in the medically ill, despite consensus that a sizeable pool of depression exists in this population. The performance of several self-rating scales in these patients is reviewed, demonstrating that satisfactory screening of depression as defined by standard case criteria can be achieved. It is proposed that it is primarily the absence of clear guidelines for treatment to be used in conjunction with these scales which will make clinicians sceptical of their value. Treatment validation of case criteria is required to demonstrate that screening for depression in medical patients is worthwhile.