Although the editors of this book are to be congratulated on their selection of experts from the USA, Australia and continental Europe, this is essentially a British book with British perspectives and intentions. This is proper, as any system of community psychiatry must operate within constraints arising from the social and historical context. Although there may be similarities in the best form for services in countries with similar structures, there will also be differences, and this makes for uncertainty in the choice of service pattern and provision: what seems to be effective in one country is not necessarily so in another. Given that, it is instructive to compare this text with the impressive volume with a similar theme edited by Douglas Bennett and Hugh Freeman more than 10 years ago (Reference Bennett and FreemanBennett & Freeman, 1991). By this test, while the principles of community psychiatry remain very similar, the project has advanced considerably. The definitions of community psychiatry proposed over the years have constantly been reframed to accommodate changing practice on the ground. And the practice is continually modified by competing and complementary value systems, as set out by Thornicroft & Szmukler in their opening chapter.

The book is large: 557 pages of text, 46 chapters. The first section covers the background of the subject in terms of social policy, methodology and epidemiology. The second and longest section describes the service system in exhaustive detail, moving from the practical application of principles through the integration of service components to the components themselves. It is this, of course, that is characteristic of community psychiatry — the fact that it relies on diverse elements acting in concert, hopefully in synergy. In Britain, although in many areas services are still emerging from an age of crass underresourcing, implementation is now sufficient for there to be serious point to debates about the right combination of provision.

One of the major differences over the past 10 years has been the increased reliance on the idea of evidence-based psychiatry. Thornicroft & Szmukler include four useful chapters devoted to the scientific background of community psychiatry. There are, of course, particular difficulties in deciding best practice in community psychiatry, and these chapters make this very clear. Likewise, in the first section of the book is a set of chapters quantifying both the extent and the impact of psychiatric disorders.

The meat of the book is the large number of chapters devoted to aspects of the service system, both its components and the way they meld together. Some of these chapters provide a clear, evidential basis for the choice of service structures, but in others it is apparent that the underlying research is much thinner, usually because it is much more difficult to carry out.

Other chapters point to inherent ambiguities in community psychiatry. For example, Rosen & Barfoot highlight the difficulty of integrating appropriate day care and sheltered work into modern forms of community psychiatry. The book certainly gives a reasonably up-to-date review of the considerations involved in continuing development of community psychiatric care.

The chapters are mainly authoritative, although some are shorter and more desultory than they need to be. Nevertheless, the editors are to be complimented on an impressive effort. It is certainly useful to anyone involved in the field of community psychiatry. Most clinicians buy relatively few books. However, psychiatry in Britain has a strong community thread, and this book is a useful access point to a very large literature. For this reason, I recommend private as well as library purchase.