Background: Relapse following abstinence from substance use is common. In this study, we evaluated the effectiveness of training, based on a cognitive-behavioral relapse prevention model in individuals with opiate dependence. Methods: A total of 50 substance abusers were selected via simple random sampling among the clients of Aftab Center, Tehran, Iran for substance dependence treatment. The subjects were randomly divided into 2 groups: routine addiction treatment program and the developed intervention, plus the routine addiction treatment program. All the subjects completed the questionnaire on demographic variables and addiction history, Inventory of drug-taking situations, and drug avoidance self-efficacy scale at baseline and after the intervention. The patients were followed-up for 3 and 6 months after the intervention. ANOVA, Chi square test, and ANCOVA test were performed for data analysis, using SPSS version 19. Results: In this study, the mean age of the subjects was 38 11. 26 years. Based on the results, social pressure was the most frequent high-risk situation, followed by negative emotional and physical states. In the 3-month follow-up, 8% and 24% of the subjects from the intervention and control groups dropped out, respectively. Nevertheless, no significant differencewas found between the groups in terms of the dropout rate. At the end of 6 months, 84% and 48% of the subjects from the intervention and control groups showed abstinence from drug use, respectively. The findings indicated a significant difference regarding self-efficacy between the groups (P < 0. 001). Conclusions: Design of interventions based on cognitive-behavioral models, which can improve self-efficacy, may be effective in relapse prevention.