Predictors of either detoxification success or failure were evaluated during an inpatient trial that compared the efficacy of methadone, clonidine and guanfacine for rapid heroin detoxification. The analysis of such predictors was stimulated by the fact that in order to achieve 90 patients who completed the study (30 in each group), a total of 170 patients had to be included. Of 80 detoxification failures, 10 occurred in the methadone group, 32 in the guanfacine group, and 38 in the clonidine group. Voluntary request for discontinuation of the detoxification schedule was the first cause of failure. There were not statistically significant differences with regard to sociodemographic characteristics and pattern of drug consumption among patients in the three groups who completed detoxification with success or failure. The treatment drug, the type of schedule and the score obtained from the Symptom Checklist-90/Revised (SCL-90/R) were the only predictors of either detoxification success or failure. Inpatient opioid detoxification would be a useful strategy for patients with more severe psychological symptoms.