The Johnson Intervention is a therapeutic technique in which members of a person's social network confront him or her about the damage that drinking and drug use has caused and the action they will take if he or she does not enter treatment. Previously, we evaluated the effectiveness of the Johnson Intervention at engaging and retaining clients in outpatient alcohol and other drug (AOD) treatment by comparing it to four other methods of referral. Although individuals who undergo a Johnson Intervention are most likely to enter treatment, the power of the Johnson Intervention to retain clients deteriorates over the course of treatment, as indicated by their diminished likelihood of completing. Given that abstinence from alcohol and other drugs is one of the first expectations placed on clients, we compared the Johnson Intervention to the other types of referral to evaluate the role of relapse during treatment. In a secondary analysis of 210 cases, we found relapse rates across the five types of referral ranged from 38% to 79%. Those in the Johnson Intervention group were more likely to relapse than three of the four other groups and, across all groups, those who relapsed were less likely to complete treatment. In spite of its high relapse rate, the Johnson Intervention is very effective in retaining those who relapse because it is very effective retaining all clients, whether they relapse or not.