Our existing guide to building a relapse prevention plan covers how to put a written plan together. This piece goes a level deeper into the psychology underneath that plan: the actual research-based categories of triggers, and why some types carry meaningfully higher relapse risk than others.
Most current thinking about triggers traces back to psychologists Alan Marlatt and Judith Gordon's cognitive-behavioral relapse prevention model, developed in the 1980s. Their core insight was that relapse is not a single, sudden decision but a process that unfolds over time, beginning with exposure to what they called a high-risk situation. Marlatt developed a detailed taxonomy classifying these situations into specific categories, and later research replicating and extending that taxonomy has generally confirmed its core structure, even while refining some of the specifics.
According to Marlatt's original research, negative emotional states, including anger, anxiety, depression, frustration, and boredom, are associated with the highest rate of relapse of any single category. These are called intrapersonal high-risk situations because they originate from within the person rather than from an external event, though they are often triggered by something happening around them, feeling bored and alone after coming home to an empty house, for example, or feeling anxious about a specific upcoming event.
The widely used HALT framework, Hungry, Angry, Lonely, Tired, captures a specific, practical subset of this category. HALT became popular in recovery communities starting in the 1980s and 1990s precisely because it gives people a fast, memorable way to check in with themselves before a craving escalates into something harder to manage. Feeling any one of these four states does not mean relapse is imminent, but it is a signal worth taking seriously and addressing directly, through food, rest, connection, or simply naming the anger, rather than ignoring it.
Situations involving another person or group, particularly interpersonal conflict such as an argument with a family member, form a second major category. These triggers are distinct from purely internal emotional states because they involve an active, external social dynamic, direct pressure to use from someone else, a confrontation, or simply being around people connected to past substance use. Marlatt's research found this category carries meaningful relapse risk, though generally somewhat lower than the intrapersonal category, and it responds to different coping strategies, refusal skills, communication techniques, and, when necessary, structured distance from specific people.
This category is worth connecting directly to our guide on stigma and language, since how family and friends communicate during a conflict can itself function as a trigger independent of the conflict's underlying content. A disagreement handled with calm, direct communication carries meaningfully different relapse risk than the same underlying disagreement handled with blame, comparison, or shaming language, even when the topic and the people involved are otherwise identical.
The third broad category covers external circumstances not primarily driven by another person: financial stress, being in a specific physical location associated with past use, exposure to cues like the sight or smell of a substance, or simply unstructured time with nothing planned. This is the category most people intuitively think of first when asked about triggers, places and things, even though the research suggests it is not actually the highest-risk category on its own.
| Category | Example | Relative Relapse Risk (Per Marlatt's Research) |
|---|
| Intrapersonal (emotional states) | Boredom, anger, anxiety, depression | Highest single-category risk |
| Interpersonal (social conflict) | Argument with a family member, direct pressure to use | Significant, generally somewhat lower than intrapersonal |
| Environmental/situational | Financial stress, specific locations, unstructured time | Meaningful, but often overestimated relative to emotional triggers |
Later research, including a multisite study called the Relapse Replication and Extension Project, tested and refined Marlatt's original taxonomy using more detailed sub-categories within each of these three broad levels, allowing clinicians to identify a person's specific pattern with more precision than the broad categories alone provide. This research effort also examined how reliably different clinicians could categorize the same relapse episode using Marlatt's system, an important question for any clinical framework meant to guide real treatment decisions, and generally found reasonable consistency in how raters classified high-risk situations, lending further support to the taxonomy's practical clinical usefulness.
It is worth noting that this system was never intended as a rigid, mutually exclusive filing system. A single difficult moment often involves more than one category simultaneously, an argument with a partner (interpersonal) that leaves someone feeling angry and alone for the rest of the evening (intrapersonal), while also happening to occur at a location associated with past use (environmental). Understanding the categories individually helps clarify what is actually happening in a specific high-risk moment, but real relapse triggers frequently layer multiple categories together rather than presenting as a single, clean type.
A common but limited approach to trigger identification is a flat list: "people, places, and things." What the research actually shows is more specific and more useful clinically: emotional states, not just external circumstances, carry the single highest relapse risk, which means a relapse prevention plan focused only on avoiding certain people or locations, while ignoring the internal emotional patterns captured by HALT, is missing the category most strongly associated with actual relapse episodes.
Consider someone who has carefully planned to avoid a specific bar and a specific old friend, both genuine environmental and interpersonal triggers. Three months into recovery, they experience a difficult, boring Sunday afternoon alone, no argument, no bar, no old friend anywhere nearby, and find themselves surprised by an intense craving. This is precisely the pattern the research predicts: the intrapersonal category, boredom and loneliness in this case, often produces stronger, less anticipated risk than the more obvious external triggers someone has already planned around. A relapse prevention plan built only around avoiding people and places would have left this specific, high-risk moment completely unaddressed.
- Build coping strategies specifically for emotional states, using HALT as a starting point, not just a plan for avoiding specific people, places, or things
- Track your own patterns honestly; the specific triggers that carry the most risk for you personally may not match the general research averages exactly
- Treat unstructured, unplanned time as its own risk category worth actively managing, not simply the absence of other triggers
- Remember that noticing a trigger is not the same as relapsing; what matters most, per the research, is your coping response in that moment
Our guide to building a relapse prevention plan covers how to translate this understanding into a concrete, written document, and our guide to what to do after a return to substance use is a direct companion if a trigger has already led to a lapse.
Frequently asked questions
Intrapersonal (emotional states like anger, anxiety, or boredom), interpersonal (conflict or social pressure from others), and environmental or situational (financial stress, specific locations, unstructured time).
Negative emotional states, called intrapersonal high-risk situations in Marlatt's research, are associated with the highest single-category relapse rate, higher than interpersonal conflict or environmental cues.
Hungry, Angry, Lonely, Tired. It's a widely used framework in recovery for quickly checking in on emotional and physical states that commonly precede cravings.
Psychologists Alan Marlatt and Judith Gordon developed a cognitive-behavioral relapse prevention model in the 1980s, including a detailed taxonomy of high-risk situations that later research has largely confirmed and refined.
No. According to the research, a high-risk situation is the trigger, but a person's coping response to that situation determines whether it leads to an actual return to substance use.
Because it overlooks the category the research shows carries the highest risk: internal emotional states like boredom, anger, and loneliness, which can trigger cravings even with no external trigger present at all.
This article was reviewed by the RehabsNearMe Editorial Team for accuracy, clarity, and relevance. Information may be sourced from publicly available treatment resources, government agencies, and healthcare references where applicable.
Last reviewed: September 2026
Related articles
Resources
