“…It was in this context that the Institute of M edicine of the US National Academ y of Science undertook their 4-year investigation (Institute of Medicine, 1990), stimulated by a view that the alcohol ® eld ª needed to prove itselfº (Saunders, 1991) in terms of treatment ef® cacy, relevance to m odern health service provision and cost-effectiveness. Peele (1990) and Miller & Sanchez-C raig (1996) have critically exam ined claims for high success rates for alcoholism with programm es making claims of abstinence rates of 60± 90%. W hile it is clear that treatm ent is more effective than no treatment (McLellan et al, 1996), and that patients who receive treatm ent in well-designed treatm ent studies are signi® cantly improved in a variety of problem severity areas at follow-up (M cLellan et al, 1996;Project M atch Research Group, 1997) the nature and amount of change is typically more modest.…”