To the Editor.— We wish to correct a misleading inference drawn from our data by Dr Maltzman 1 in his recent letter. From a prepublication manuscript, Dr Maltzman drew the figure that only 10% of our clinical sample were engaging in controlled and asymptomatic drinking at long-term follow-up. This figure is correct, but is erroneously contrasted with a recovery rate of 60% to 65% that he describes as "the industry norm for abstinence-oriented programs." This quoted figure, based on a comment made in an interview, is a gross and unsubstantiated estimate of outcomes among "those who finish treatment" and is quite at variance with empirical evidence from properly conducted clinical trials. In a review restricted to outcome studies with at least one year of follow-up data, Costello 2 concluded that the average success rate (abstinent plus improved cases) after alcoholism treatment is 26%. A similar picture emerged from Vaillant's 3