The basic design of a randomised clinical trial seems attractively simple. Patients are allocated at random to one of two or more interventions, the groups are then followed up and the outcomes are compared. Underpinning this basic design, of course, are a large number of complex decisions which investigators grapple with during the course of protocol development. Fortunately, the area has produced more than its fair share of classic textbooks over the years. Some of these achieve such a standard on first publication that they maintain their status as beacons of good judgement to guide researchers as they design their trials, as well as those critically appraising trials. Among these are Stuart Pocock's Clinical Trials: A Practical Approach (Wiley, 1984) and Curtis Meinert's Clinical Trials: Design Conduct and Analysis (Oxford University Press, 1986) which have both remained in print without new editions for more than two decades.

Everitt & Wessely's Clinical Trials in Psychiatry was recognised to be an excellent introduction on first publication in 2003, giving both a readable and authoritative overview as well as a special focus on the particular practical issues and difficulties that occur in trials in psychiatry. The 5 years since the first edition have seen major advances in methodology and the authors have updated the book accordingly. In particular, more attention is paid to the distinction between the highly controlled and intensive phase 3, or explanatory trials, and the more pragmatic effectiveness trials that have increasingly been reported over the past decade. The authors highlight the trade-offs involved in designing a trial in one way or another: no single trial can answer all the relevant questions. As in the first edition, there is a valuable chapter on statistical issues that are particularly relevant in psychiatry such as dealing with repeated longitudinal outcome measures.

Clinical Trials in Psychiatry covers all the most important issues and will be useful to all clinicians who are involved in conducting, or using the results of, clinical trials. It provides a highly persuasive account of the unique scientific advantages of randomised trials for those who remain unconvinced. It is very readable and even the very occasional errors are amusing (the reference to military tuberculosis, for example on page 21).

A book to read from beginning to end and then place next to Pocock and Meinert.