The author focuses on therapeutic communities, particularly the KEY and CREST programs, applied in the state of Delaware since 1988. There programs are aimed at substance users living either in correctional settings (KEY) or on probation (CREST). This research tries to evaluate both programs according to the subject's relapse rate after 6 months. Based on the results, it has been determined that the subjects who participated in these programs successively had better chances of never using drugs again than those who had been in only one of the two programs, or neither. Other American experiments of the same kind have shown similar results.