1. The programme at a glance
Indication. Mild to severe cannabis use disorder, in adults in outpatient care, and in adolescents with their family taking part, once the risks in section 11 have been looked for and those who need medical care have been referred.
Reference model. A cognitive behavioural therapy of the addictions: motivational interviewing at the start (section 32), functional analysis of the situations in which use occurs (section 6), skill acquisition — getting through withdrawal, craving, refusal, emotions without the product, sleep — and relapse prevention. Contingency management, which gives concrete rewards for abstinence verified by assays (section 33), is added wherever it is available. In adolescents, the same work sits inside a family approach (section 34).
Format. For adults, twelve weekly 50-minute sessions, then three booster sessions at one month, three months and six months (section 31). For adolescents, twelve meetings over three to four months, six of them with the young person alone, two with the parents alone and four joint sessions (section 34). The number of sessions is part of the protocol: interventions of more than four sessions spread over more than a month do better than shorter formats, at least in the short term (Gates et al., 2016).
Mechanism targeted. Replacing a response that has become automatic with a chosen response, in situations identified in advance, and replacing what the product used to do — getting you to sleep, calming you down, filling time — with something else. Willpower is not what is worked on: what is worked on is the hours, the places, the nights, and the hour that follows a first joint smoked again after stopping (session 11).
| Session |
Focus |
Session deliverable |
| 1 |
Assessing and looking for risks |
Risks checked, record started |
| 2 |
Ambivalence and the goal |
Goal written down, review date |
| 3 |
The record and the situations |
Four situations taken apart |
| 4 |
Preparing the quit date |
Date set, environment cleared |
| 5 |
Getting through withdrawal |
Withdrawal rated day by day |
| 6 |
Craving |
One craving ridden out in session |
| 7 |
Refusing, and the network |
Three refusals tried, contacts sorted |
| 8 |
Sleeping without cannabis |
Sleep rules in place |
| 9 |
What cannabis was solving |
One written answer per function |
| 10 |
High-risk situations |
Written plan for five situations |
| 11 |
The lapse |
What to do written before it happens |
| 12 |
Review, relapse, maintenance |
Measures repeated, maintenance plan |
What the person takes away. Nine printable worksheets, listed in section 40: my record, my balance sheet and my goal, my situations, my first fifteen days, my craving, my refusals and my network, my plan for what comes next, our agreement — for the adolescent and their parents —, and the corner for relatives and parents.
What sets this programme apart from the other manuals on this site. Three things. A risk section placed before everything else, which refers on to other programmes when a psychosis or a suicide risk appears. A goal discussed without dogma, with the honesty to say that lasting abstinence is a minority outcome. And a full adolescent and family strand, with its own rules of confidentiality.
2. Before you start
The four decisions to take first
Is there a risk that changes what you do? Psychotic symptoms, suicidal thoughts, cyclical vomiting, pregnancy, driving while affected, synthetic cannabinoids. Section 11.
What is being used, and how? This is no longer the same product as twenty years ago. Section 4.
What is the goal? Stopping or cutting down. Section 17: the question arises at session 2 and it is reviewed later.
And who is in the room? An adult alone, an adult and a relative, or an adolescent and their family. For an adolescent, confidentiality is settled in writing at the first meeting, before anything at all is said. Sections 34 and 35.
What this programme does not treat
A psychotic episode. It spots it, it refers on, and it points to First-episode psychosis: a therapist's manual; work on use then continues within the service treating the psychosis, not alongside it.
Emergencies. Severe intoxication, poisoning in a child, uncontrollable vomiting with dehydration, a suicidal crisis, seizures after synthetic cannabinoids.
Severe polysubstance use, which belongs in an addiction service; for alcohol, see Alcohol use disorder: a therapist's manual.
And prescribed medical cannabis. Section 12.
How to use it
Read the whole thing before the first session, in particular sections 8, 11, 17, 32 and 34: withdrawal, the risks, the goal, motivational interviewing, and the adolescent strand. Every session follows the same frame: the aim, the steps, what you say, the common mistakes, and the criterion for moving on.
Two warnings specific to this presenting problem.
Playing it down and dramatising it lose the person just as surely as each other. The patient often knows the product better than you do, and switches off at the first exaggerated figure. They also switch off when their daily use is treated as a matter of lifestyle.
Ambivalence is not bad faith. Cannabis renders real services — it sends you to sleep, it calms you, it fills time, it brings people together. A patient who tells you so is handing you the material for session 2.
3. Four pictures not to be confused
1. Occasional use without a disorder
What you observe. Episodic use, without loss of control, without anything appearing on stopping, without notable repercussions.
What that implies. Honest, brief information — driving, pregnancy, vulnerability to psychosis, pointers drawn from the international guidelines on lower-risk use (section 17). This programme is not indicated.
2. Regular use at risk
What you observe. Frequent use, sometimes daily, without clear loss of control, but with risk factors: a young age, very concentrated products, driving, a pregnancy, a family history of psychosis.
What that implies. A brief intervention: measure the use, feed the figures back to the person (the personalised feedback described in section 32), set a numerical goal with them, and see them again to take stock. Two well-conducted sessions did better than a waiting list in dependent adults (Marijuana Treatment Project Research Group, 2004); using them at this earlier stage is a reasonable extrapolation. And monitoring of the trajectory, because dependence most often sets in during the first years of use (Lopez-Quintero et al., 2011).
3. The use disorder
What you observe. A loss of control, attempts to cut down that fail, a growing place taken by the product, repercussions at school, at work or in relationships, and often withdrawal on stopping.
What that implies. The programme described here.
4. Use in the service of another disorder
What you observe. Cannabis is used to sleep, to calm a social anxiety, to switch off re-experiencing, to soothe a long-standing restlessness, to bear a collapsed mood.
What that implies. Both are treated, in an order that depends on severity. Massive daily use makes it impossible to treat anything else; moderate use in the service of an anxiety is treated along with it.
The three sorting questions
"What happens when you do not smoke for three days?" It isolates withdrawal, one of the signs that the body has adapted.
"What is the joint for?" It isolates the function, which decides what will have to be put in the product's place (session 9).
"Has it ever happened that you felt very frightened, felt watched, or heard things, while smoking or afterwards?" It opens the question of psychosis, and it is asked of everyone.
4. Today's products
You cannot assess someone's use without knowing what is being used, and the market has changed a great deal. Someone who says "I smoke the way I did at twenty" is not smoking the same product.
Herbal cannabis and resin. A meta-analysis of studies that tracked the content of cannabis samples in seven countries shows a rise in THC content (delta-9-tetrahydrocannabinol, the main psychoactive compound in cannabis, responsible for the high) between 1970 and 2017, more marked in resin than in herbal cannabis, while cannabidiol (CBD) content has stayed stable. In herbal cannabis, the rise is mainly due to the place taken by sinsemilla varieties, which are much more concentrated (Freeman et al., 2021). Grams no longer say much about the dose.
Concentrated extracts and vaping. They deliver a great deal of THC in a short time: inhaling highly concentrated extracts can produce unwanted acute effects, and the higher the THC content, the greater the risks (Fischer et al., 2022). And a vaping liquid sold outside a controlled supply chain may contain something other than what the label announces: in the United States, products sold as CBD, most of them vaped, in fact contained a synthetic cannabinoid and no CBD (Horth et al., 2018).
Edible products. The effect comes late and lasts a long time; the trap is taking another dose before the first has worked. Left within reach, they cause accidental poisoning in young children, which is an emergency: the question arises for every parent who uses.
Synthetic cannabinoids. Laboratory molecules that act on the same receptors as THC, often far more potent and of unpredictable composition (Cooper, 2016), sometimes concealed in products presented as CBD (Horth et al., 2018). A systematic review of the cases seen in hospital, in emergency departments and in poison centres describes mainly tachycardia, agitation and nausea, but also, more rarely, cardiovascular events, acute kidney injury, seizures, psychotic states and deaths (Tait et al., 2016). Routine assays generally do not allow their use to be confirmed (Cooper, 2016), and the World Health Organization's classification, ICD-11, gives them a separate category (section 7).
Semi-synthetic cannabinoids. Derivatives sold for a time openly as a "legal" alternative to cannabis, on which little data is available (Ujváry et al., 2023), and whose status keeps changing: the first of them, hexahydrocannabinol (HHC), was placed under international control in 2025 (World Health Organization, 2025). In session, what counts is not their name: it is what the person felt from them, and how much of them they take.
CBD. It does not produce the intoxication of THC (Connor et al., 2021), and it is not an approved treatment for the disorder (section 16). The labelling of over-the-counter products is unreliable: in a study of products sold online, fewer than a third were correctly labelled, and one in five contained THC, sometimes in a quantity sufficient to produce intoxication or impairment (Bonn-Miller et al., 2017). Its drug interactions are a matter for the doctor.
5. What keeps the disorder going
Immediate reinforcement. The product relaxes, sends you to sleep, anaesthetises an emotion, within minutes; the drawbacks are deferred. So the work is on what happens in the ten minutes after taking it, not on the risks ten years on.
Tolerance and the reversal of function. With daily use, the sought-after effect wears thin and use serves more and more to keep from feeling bad: "I no longer smoke to get high, I smoke to feel normal".
Withdrawal and the relief of starting again. The irritability, the insomnia and the anxiety of the first few days are relieved immediately by a dose. The person concludes that cannabis is indispensable to them, when what they are observing is the effects of stopping. It is the most underestimated mechanism, and it is defused by describing it beforehand. Section 8.
Cues. A time of day, a place, the ritual of preparation, a smell, a screen, certain friends. Each triggers a craving without anything having been decided.
The environment. A product that is available, a dealer reachable in one message, friends who use. Changing it is not cheating: it is one of the most useful measures there is, and the least valued.
Function. Sleeping, calming down, filling time, switching off an anger, easing a pain. Until it is replaced, the product comes back.
Identity. Smoking is sometimes part of who someone is. Stopping then raises one more question — who am I without it — which is worked on at session 7.
Tobacco. In those who mix cannabis with tobacco, two dependences sit on top of each other: stopping one wakes the craving for the other. The question is dealt with at session 4.
The full circle
A situation triggers a craving. The dose relieves it immediately. The consequences come later — a restless night, a hazy memory, lateness, an argument, guilt. The anxiety and the insomnia that follow become triggers in their turn. And the next dose comes earlier in the day.
This is the diagram you draw at session 3.
Frequently asked questions
Does cannabis really cause dependence?
Yes, in a minority of users: a little under one person in ten among those who have used it, in a large North American survey (Lopez-Quintero et al., 2011), and far more with daily use, above all when started young (Connor et al., 2021; Fischer et al., 2022). It is recognised by the loss of control and by the place taken, not by the quantity. Section 7.
Is withdrawal dangerous?
Not in itself: it causes neither seizures nor delirium, unlike alcohol withdrawal. It is, on the other hand, common, unpleasant, and it makes people start again. Medical advice is essential where there is polysubstance use or synthetic cannabinoids. Section 8.
The patient only wants to cut down.
That is a legitimate goal in an adult who presents none of the situations in which stopping is essential: psychotic symptoms or bipolar disorder, hyperemesis, pregnancy, breastfeeding or planning a pregnancy, synthetic cannabinoids, driving while affected that does not yield to rules or a safety-critical job, repeated failures of reduction. Put figures on it, write it down, set a review date, and say frankly if they are one of the cases in which reduction is not reasonable. Section 17.
Can CBD help someone to stop?
The data do not allow it to be recommended (Spiga et al., 2025), and the labelling of over-the-counter products is unreliable (Bonn-Miller et al., 2017). No medication is approved in this indication. Section 16.
He had persecutory ideas after a night out.
Ask the questions in section 11, refer promptly to a doctor, and make stopping the goal. In a large cohort, nearly half of the cannabis-induced psychoses were followed by a diagnosis of schizophrenia or bipolar disorder (Starzer et al., 2018). See First-episode psychosis: a therapist's manual.
She has been vomiting for months and nobody can find anything.
Ask whether she uses cannabis regularly and whether hot showers relieve her. Cannabinoid hyperemesis syndrome is often diagnosed late; it is a matter for the doctor, and stopping emerges as the best treatment (Sorensen et al., 2017). Section 11.
The parents want to know whether their son is still smoking.
You will tell them whether he comes, how the work is going, and you will alert them if he is in danger. The detail of his use stays between him and you, unless there is danger: that is the condition for his saying anything useful. Section 35.
The parents want to do urine tests at home.
Advise against them outside an agreed care framework: they often damage the relationship, and their effect on use has not been demonstrated. If a check is useful, it belongs within a protocol with a doctor, known to and accepted by the young person. Sections 33 and 34.
The adolescent refuses to have his parents involved.
Begin with the individual work, which has evidence behind it, and go on looking for a place, however small, for a trusted adult. The refusal often changes once the young person has seen that confidentiality is kept.
He smoked again and has not come for a month.
Write or call, without reproach, offering an appointment. Clinicians observe that it is often shame, not indifference, that keeps people away. Section 29.
Should tobacco be stopped at the same time?
The question is not settled. What matters is that it is put and decided before the quit date, and that help with stopping smoking is offered if the person stops both. Session 4.