With the introductory chapters out of the way this book can settle to the serious work in hand. Chapter 4, by Regier et al., is an excellent start. It takes a 50-year perspective on the delivery of mental health services. It outlines the flurry of training that followed the Stirling County and Mid-Town Manhattan studies, the first studies that demonstrated the importance of non-psychotic mental disorders. Treatment capacity increased, but the proportion of cases in treatment did not, as revealed by the Epidemiologic Catchment Area (ECA) and the 1990 National Comorbidity Survey (NCS) studies. It notes the changes produced by legislation and the changes brought about by managed care. The message is of support for good data on which good service plans can be built. Kessler (Chapter 5) takes the argument one step further. He says that the problem of unmet need for people with mental disorders is part of the larger problem of resource allocation in health, and, if we are to succeed in reducing the unmet need, we will have to gather immaculate data to show that the health gains in psychiatry, in human capital terms, are sufficiently great to warrant increases in funding in competition with the costs and benefits of interventions in the physical diseases. The examples he finds as to how this could be done are quite exciting. Chapter 8, by Rupp and Lapsley, reinforces this perspective.