Relapse prevention
(RP)
- Method
- Moderate evidence
- Published October 19, 2026
- 8 sources
- 5 min read
In short
Related in this topic
On this page
Key facts
- Used in
- Cognitive behavioral therapy
- Developed by
- G. Alan Marlatt and Judith R. Gordon; dynamic reformulation by Katie Witkiewitz and G. Alan Marlatt
- Origin
- Model set out by Marlatt and Gordon in 1985; reformulated as a dynamic model in 2004
- Also known as
- Relapse prevention model; Cognitive-behavioral model of relapse; Marlatt’s relapse prevention model
State of the evidence
Moderate evidenceA meta-analysis of 26 studies with 9,504 participants found relapse prevention generally effective, particularly for alcohol problems, with the largest effects in uncontrolled comparisons assessed soon after treatment[1]. Reviews find it no more effective than other active treatments in producing abstinence[2], and a meta-analysis of cognitive-behavioral therapy for alcohol and drug use disorders, which incorporates relapse prevention, found no advantage over other specific therapies[3][4]. A mindfulness-based version has mixed trial results[5][6].
Evidence labels summarise the research literature; they are not recommendations. How labels are assigned
Relapse prevention (RP) is a cognitive-behavioral approach to maintaining a change in an addictive behavior. Developed by G. Alan Marlatt and Judith Gordon, it combines a model of how relapse comes about with a set of cognitive and behavioral strategies intended to prevent or limit relapse episodes[2][7]. Marlatt’s colleagues contrast it with approaches that treat relapse as an end state equivalent to treatment failure: in the RP model, relapse is a process that begins before the first use after a period of change and continues after it[2].
The model distinguishes a lapse, an initial return to the behavior, from a relapse, a return to the former pattern or abandonment of the goal; a lapse raises the risk of relapse but does not make it inevitable[2]. Developed for addictive behaviors, RP has also become an adjunct in the treatment of other conditions, including depression, sexual offending and schizophrenia[8]. Its components are now built into most cognitive-behavioral treatments for substance use, which, according to Marlatt and colleagues, has made the model itself hard to evaluate separately[4].
Immediate determinants of relapse
The original model groups the causes of relapse into immediate determinants and covert antecedents[2]. The central immediate determinant is the high-risk situation. In Marlatt’s taxonomy, drawn from accounts of relapse by people treated for alcohol problems, the situations include negative emotional states, interpersonal conflict, social pressure, positive emotional states, exposure to cues associated with drinking and tests of personal control. Negative emotional states and interpersonal conflict accounted for more than half of the relapse episodes in his analysis[2].
Whether a high-risk situation leads to a lapse depends, in the model, on the person’s coping response. Coping successfully is assumed to raise self-efficacy, a perceived mastery over the specific situation, while low self-efficacy leaves a person feeling unable to resist[2]. Outcome expectancies add to the risk when a person anticipates the immediate positive effects of a substance, such as relief from stress, and discounts the delayed negative ones[2].
The fourth determinant is the abstinence violation effect, a reaction to a first lapse. A person who attributes the lapse to personal failure is said to experience guilt and negative emotion that can lead to further use, and one who attributes it to stable, global, internal causes, such as having no willpower, is more likely to abandon the goal than one who attributes it to difficulty coping with a particular situation[2].
Figure 1A theoretical model, not an observed sequence. In Marlatt and Gordon’s cognitive-behavioral account, an effective coping response to a high-risk situation raises self-efficacy and lowers the probability of relapse. An ineffective response lowers self-efficacy, which with positive outcome expectancies can lead to a lapse, the abstinence violation effect and an increased probability of relapse. The model of Marlatt and Gordon (1985), drawn as Larimer, Palmer and Marlatt (1999) summarize it.
Text description
A theoretical model drawn as a flow from top to bottom. At the top, a high-risk situation, such as negative emotional states, interpersonal conflict or social pressure. It branches in two. On the left, an effective coping response leads to increased self-efficacy, confidence in coping with the situation, and then to a decreased probability of relapse. On the right, an ineffective coping response leads to decreased self-efficacy, feeling unable to resist; then to a lapse, an initial return to the behavior; then to the abstinence violation effect, feelings of guilt and failure; and then to an increased probability of relapse. A separate box on the left, also contributing, shows positive outcome expectancies, anticipating the immediate positive effects of the substance. Dashed arrows lead from it to the lapse and to the increased probability of relapse.
Covert antecedents
Less obvious factors are said to set up high-risk situations. Marlatt and Gordon proposed that the covert antecedent most strongly related to relapse is the balance in a person’s life between perceived obligations, the “shoulds,” and enjoyable activities, the “wants”[2]. A run of choices that each seem inconsequential, which the model calls apparently irrelevant decisions, can bring a person into a situation of overwhelming risk while disavowing responsibility for it. Urges and cravings are treated as responses triggered by psychological or environmental stimuli, through classical conditioning to cues associated with past use and through expectations of immediate pleasure[2].
Intervention strategies
The model pairs each step of the relapse process with intervention strategies, grouped as skills training, cognitive restructuring and lifestyle balancing[2]. Specific strategies include identifying a client’s high-risk situations, through interviews and self-monitoring, and rehearsing coping responses for them in role plays; setting small, achievable goals to build self-efficacy; examining expectancies about a substance’s effects against both its immediate and delayed consequences; and preparing a plan for limiting a lapse[2]. Cognitive restructuring is used to reframe a lapse as a mistake in learning rather than a sign of personal weakness, which is meant to counter the abstinence violation effect[2].
Global strategies address the covert antecedents. They include assessing the balance of “wants” and “shoulds,” developing activities with long-term benefits to mood and coping, removing cues from the person’s surroundings, and treating urges as conditioned responses that rise and fall. One imagery technique, urge surfing, presents a craving as a wave to be watched rather than fought[2].
The dynamic reformulation
In 2004 Witkiewitz and Marlatt responded to criticisms of the original model with a reformulation that focuses on dynamic interactions among multiple risk factors and situational determinants[8]. As later summarized by Marlatt and colleagues, the reformulated model treats relapse as a complex, nonlinear process[4]. Tonic processes are stable influences that set a person’s initial threshold for relapse, such as familial risk, personality and relatively stable cognitions like global self-efficacy. Phasic responses are transient: urges, mood, momentary changes in self-efficacy or motivation, coping in the moment and the effects of a lapse itself. In this account, tonic processes determine who is vulnerable and phasic processes determine when relapse occurs[4].
The reformulation draws on nonlinear dynamical systems theory and catastrophe theory, and it predicts that a small change in one factor, such as a momentary drop in self-efficacy, can have a disproportionate effect and that relapse can be sudden. In one analysis of post-treatment drinking, catastrophe models accounted for more than twice the variance explained by linear models[4].
Meta-analytic evidence
Marlatt and colleagues describe Carroll’s review of 24 interventions as the first comprehensive review of RP outcomes. It found RP not consistently superior to other active treatments but promising for reducing the severity of relapse and for durable gains, with some effects emerging only at later follow-up[4]. A meta-analysis of 26 studies, with 70 hypothesis tests and 9,504 participants, found RP generally effective, particularly for alcohol problems[1]. The effect on substance use was small (r = 0.14), and the effect on psychosocial functioning was larger[4]. Effects were largest for alcohol and polysubstance use, when RP was combined with medication, and in uncontrolled before-and-after comparisons assessed soon after treatment[1].
Other assessments are less favorable. RP is not associated with higher abstinence rates than other credible treatments, although some trials found less drinking among people who did resume[2]. A methodological review of alcohol treatments ranked relapse prevention 35th of 46 approaches[4], and meta-analyses up to 2009 found no support for skills-based RP in preventing relapse to smoking[4]. Tests of the model itself have been mixed. A multisite replication found the taxonomy of high-risk situations only moderately reliable at its most specific level and modest in predictive validity, while coping in those situations strongly predicted outcome[2]. Because RP is embedded in broader programs, much of its recent evidence comes from trials of cognitive behavioral therapy. A meta-analysis of 30 such trials for alcohol and drug use disorders found moderate effects against minimal treatment, effects against nonspecific therapies that were significant early but not late, and no advantage over other specific therapies[3].
Mindfulness-based relapse prevention
Mindfulness-based relapse prevention (MBRP) combines RP with mindfulness practice. In contrast to the cognitive restructuring of traditional cognitive-behavioral therapy, it stresses nonjudgmental attention to thoughts and urges[4]. Interest in it followed supportive evidence for mindfulness-based interventions in preventing relapse of depression, the use studied for mindfulness-based cognitive therapy[4] . In a trial with 286 people who had completed initial treatment for substance use disorders, eight weekly group sessions of either MBRP or standard RP reduced relapse to drug use and heavy drinking at six months compared with usual aftercare. RP delayed the first drug use, and at twelve months MBRP participants reported fewer days of substance use and heavy drinking than both comparison groups. Three of the authors received payment for MBRP trainings[5]. A later meta-analysis of nine randomized trials with 901 participants found no significant difference between MBRP and comparison conditions on relapse or frequency of use, small effects on craving and on negative consequences, and evidence of low or very low quality[6].
Sources8
Numbered in order of first citation.
- 1.Irvin, J. E., Bowers, C. A., Dunn, M. E., & Wang, M. C. (1999). Efficacy of relapse prevention: A meta-analytic review. Journal of Consulting and Clinical Psychology, 67(4), 563–570. https://doi.org/10.1037/0022-006X.67.4.563
- 2.Larimer, M. E., Palmer, R. S., & Marlatt, G. A. (1999). Relapse prevention: An overview of Marlatt’s cognitive-behavioral model. Alcohol Research & Health. pmc.ncbi.nlm.nih.gov/articles/PMC6760427/
- 3.Magill, M., Ray, L., Kiluk, B., Hoadley, A., Bernstein, M., Tonigan, J. S., & Carroll, K. (2019). A meta-analysis of cognitive-behavioral therapy for alcohol or other drug use disorders: Treatment efficacy by contrast condition. Journal of Consulting and Clinical Psychology, 87(12), 1093–1105. https://doi.org/10.1037/ccp0000447
- 4.Hendershot, C. S., Witkiewitz, K., George, W. H., & Marlatt, G. A. (2011). Relapse prevention for addictive behaviors. Substance Abuse Treatment, Prevention, and Policy, 6(1), 17. https://doi.org/10.1186/1747-597X-6-17
- 5.Bowen, S., Witkiewitz, K., Clifasefi, S. L., Grow, J., Chawla, N., Hsu, S. H., Carroll, H. A., Harrop, E., Collins, S. E., Lustyk, M. K., & Larimer, M. E. (2014). Relative efficacy of mindfulness-based relapse prevention, standard relapse prevention, and treatment as usual for substance use disorders: A randomized clinical trial. JAMA Psychiatry, 71(5), 547–556. https://doi.org/10.1001/jamapsychiatry.2013.4546
- 6.Grant, S., Colaiaco, B., Motala, A., Shanman, R., Booth, M., Sorbero, M., & Hempel, S. (2017). Mindfulness-based relapse prevention for substance use disorders: A systematic review and meta-analysis. Journal of Addiction Medicine, 11(5), 386–396. https://doi.org/10.1097/ADM.0000000000000338
- 7.Marlatt, G. A., & Gordon, J. R. (1985). Relapse prevention: Maintenance strategies in the treatment of addictive behaviors. Guilford Press.
- 8.Witkiewitz, K., & Marlatt, G. A. (2004). Relapse prevention for alcohol and drug problems: That was Zen, this is Tao. American Psychologist, 59(4), 224–235. https://doi.org/10.1037/0003-066X.59.4.224
Further reading
- Marlatt, G. A., & Gordon, J. R. (1985). Relapse prevention: Maintenance strategies in the treatment of addictive behaviors. Guilford Press.
- Larimer, M. E., Palmer, R. S., & Marlatt, G. A. (1999). Relapse prevention: An overview of Marlatt’s cognitive-behavioral model. Alcohol Research & Health. pmc.ncbi.nlm.nih.gov/articles/PMC6760427/Alcohol Res Health 1999;23(2):151–160; PMID 10890810; no DOI. Full text read.
- Hendershot, C. S., Witkiewitz, K., George, W. H., & Marlatt, G. A. (2011). Relapse prevention for addictive behaviors. Substance Abuse Treatment, Prevention, and Policy, 6(1), 17. https://doi.org/10.1186/1747-597X-6-17
Related entries 12
- Self-efficacyConcept
- Cognitive behavioral therapy(CBT)Therapy
- Mindfulness-based cognitive therapy(MBCT)Therapy
- Cognitive restructuringMethod
- Classical conditioningConcept
- Motivational interviewing(MI)Method
- Community reinforcement approach(CRA)Therapy
- Cognitive-behavioral therapiesTopic
- Matrix modelTherapy
- Motivational enhancement therapy(MET)Therapy
- Seeking Safety(SS)Therapy
- Psychological treatments for substance useTopic
Cite this entry
APA 7
Psych Dictionary. (2026, October 19). Relapse prevention (RP). In Psych Dictionary. https://www.psychdictionary.com/methods/relapse-prevention
MLA 9
“Relapse Prevention (RP).” Psych Dictionary, 19 Oct. 2026, www.psychdictionary.com/methods/relapse-prevention.