The study was designed to duplicate as closely as possible the Bullock study on alcoholic subjects, yet still mesh with the standard treatment protocol of the Substance Abuse Unit. Study patients were randomly divided into a treatment group of 25 and a control group of 11 patients. The initial recruiter, patients, and independent raters were all "blinded" as to group placement; only the acupuncturist knew the difference. Acupuncture was given on a predetermined schedule to both treatment and control groups, with status assessed regularly by independent, "blinded" raters. The study failed to show a significant difference between treatment and control groups; however, the study patients, as a whole, did remain in treatment longer than a retrospectively analyzed group that received no acupuncture. There are several possible explanations for this study's failure to show significant benefits from acupuncture treatment. First, it was a small study, with only four patients who completed all three phases of treatment. Second, the study did not include a "no acupuncture" group for comparison; it is possible that mere participation in the study affected the outcome of control-group patients as well as treatment-group patients. Third, the inpatient status of the study group necessarily affected patients' answers on the reporting scales. Finally, because of the inpatient setting, the group setting for the treatments, and the impossibility of monitoring patients' interaction among themselves, there is some doubt about whether patients remained ignorant of their designation throughout the course of the study. 10 references