With the advent of effective physical therapies, psychiatrists have become increasingly interested in the problems posed by clinical trials. Professor Bradford Hill, in the Preface to the Sixth Edition of his Principles of Medical Statistics (1) noted that “… passage of time has … brought clinical trials into prominence and fashion …”. More recently Professor Pickering has described the clinical trial as “the most important new method introduced into clinical science in recent years” (2). Some interesting differences between controlled and uncontrolled trials are well brought out in the paper by Foulds (3) in which he found that therapeutic success was claimed in 85 per cent. of studies not using control groups. Among those studies using control groups, however, only 25 per cent. obtained a positive result. He concluded that “claims for the success of a treatment are closely associated with absence of the means whereby these claims can be scientifically substantiated”.