1. The programme at a glance
Indication. An adult who takes a benzodiazepine or a related drug every day, or nearly every day, and has done so for at least several weeks, to sleep or to ease their anxiety, and for whom stopping has been decided with their doctor, or who wishes to stop and accepts that the taper will be conducted by the doctor. The related drugs, often called Z-drugs, are sleeping pills that do not have the chemical structure of benzodiazepines but act on the same receptors. The French Haute Autorité de santé (2015) and NICE, the British institute that sets standards of good clinical practice (National Institute for Health and Care Excellence, 2022), deal with the two families together, as this manual does; the joint US guideline covers benzodiazepines alone, while noting that the related drugs carry comparable risks (Brunner et al., 2025). The indication covers in particular use that has gone on for years although the original reason has long since faded, previous attempts to stop that have failed, and the older person (section 30).
Reference model. Cognitive behavioural therapy conducted in parallel with a gradual taper supervised by the prescriber. It is the best-supported combination: adding a psychological intervention to gradual tapering increases the proportion of people who stop, compared with tapering alone, at least in the first months (Parr et al., 2009; Darker et al., 2015). The joint US guideline asks for this support to be offered (Brunner et al., 2025), and NICE for group cognitive behavioural therapy to be considered (National Institute for Health and Care Excellence, 2022). The programme borrows from three sources: motivational interviewing for the decision, cognitive behavioural therapy for insomnia and for anxiety to replace what the pill used to do, and relapse prevention for what comes afterwards.
Format. Ten individual 60-minute sessions, the first lasting 90 minutes, in three phases. Sessions 1 to 4 prepare for stopping, once a week, before the first reduction. Sessions 5 to 8 accompany the taper: each takes place in the days following a step, that is, a stage of reduction set by the prescriber. Sessions 9 and 10 cover the weeks following the last dose. Three booster sessions follow, at one, three and six months after session 10. Section 14 sets out this timetable in detail and says what it rests on.
Mechanism targeted. Three changes, in this order. That the person decides to stop with full knowledge of the facts, knowing what the medication still gives them and what it costs them. That they have other ways of sleeping and of coping with anxiety before the dose goes down. And that they learn to read the symptoms that accompany the taper as the temporary adjustment of a body that has grown used to the drug, and not as proof that they need the pill.
| Session |
Focus |
Session output |
| 1 |
Assess and make safe |
Risks sorted, baseline measures, record handed over |
| 2 |
Understand and decide |
Written balance sheet, decision made, letter to the prescriber |
| 3 |
Sleeping differently |
Fixed rising time and bed rules written down |
| 4 |
The tapering plan |
Prescriber's timetable copied over, emergency signs known |
| 5 |
Anxiety without a pill |
Two tools practised in session |
| 6 |
Reading the symptoms of the taper |
Four questions applied to a real week |
| 7 |
The pill, the urge and the habits |
A plan for each situation, pill left at home |
| 8 |
The last steps |
Core belief put to the test, ending prepared |
| 9 |
The first weeks without |
Leftover packs returned, what to do after a slip written down |
| 10 |
Review and plan for what comes next |
Measures repeated, written plan, boosters dated |
What the person takes away. Eight printable worksheets, listed in section 35: my record, my balance sheet, my tapering plan, my nights, when it rises, my habits and my urges, the corner for those close to you, my plan for what comes next.
What distinguishes this programme from the other manuals on this site. Several manuals already address stopping a sleeping pill or an anxiolytic, but as one stage of a treatment that is aimed at something else: Treating chronic insomnia in adults: a therapist's manual, in its session devoted to sleeping pills and its section on coming off sleeping pills; Treating generalised anxiety disorder in adults: a therapist's manual, regarding the patient on long-term benzodiazepines; Panic disorder and agoraphobia: a therapist's manual, in its section on the place of medication and its session devoted to medication and daily life. When the original disorder is active, those manuals remain the indication. This one is designed for the opposite situation, in which it is the medication itself that has become the problem: its sessions follow the prescriber's timetable, and it devotes a whole session to distinguishing withdrawal, rebound and relapse, because confusing them often leads people to go back to their previous dose (section 4).
2. Before you begin
The four preliminary decisions
Has stopping been decided with the prescriber, or can it be? That is the condition of the programme. The psychologist does not conduct a taper: they accompany it. If the person comes alone, session 2 prepares a letter to the doctor with them, and session 4 only takes place once the doctor has drawn up the tapering plan.
Is there a risk that calls for specialist medical care? Very high doses, alcohol dependence, a history of complicated withdrawal or a severe psychiatric disorder call for joint specialist care, sometimes for inpatient withdrawal, and an associated opioid treatment calls for close coordination between the doctors (section 9).
Is this physical dependence at the prescribed dose, or a use disorder? The first is the indication of the programme; the second falls first to addiction services, with this programme in support (section 3).
Is the disorder for which the medication was prescribed still active? Untreated chronic insomnia, generalised anxiety, panic disorder or depression is treated first or at the same time, according to its own manual (section 8). Stopping the pill without putting anything in its place, in a person whose disorder is still there, paves the way for going back to it.
What this programme does not treat
It does not conduct the withdrawal. The speed of the taper, the choice of steps, any switch of medication before the taper and what to do in the event of strong symptoms belong to the prescriber (section 12).
It does not treat a use disorder on its own. That falls first to an addiction service, which this programme can accompany without replacing it (sections 3 and 30).
How to use it
Read the whole thing before the first session, in particular sections 4, 9, 12 and 27: the definitions of withdrawal, rebound and relapse, safety, what belongs to the prescriber, and how to read the symptoms of the taper.
Each session is described on the same framework: the aim, the sequence of steps, what you say, what you do not say where that is useful, the common errors, and the criterion for moving on.
Two pitfalls specific to this presenting problem should be kept in mind from the outset. The first is talking about doses: the person will ask you how much to cut, and when, and your answer is always that this is decided with the prescriber. The second is moralising: physical dependence is the expected consequence of regular use (Brunner et al., 2025), and speaking of addiction or weakness would be inaccurate in most cases. Section 14 describes the therapist's stance that avoids both.
3. Four situations not to be confused
People who take a benzodiazepine or a related sleeping pill are in one of four situations. They do not call for the same approach.
1. Short-term or occasional use
What you observe. A treatment lasting a few days or a few weeks, or irregular use, a few evenings a month.
What points to it. The absence of prolonged daily use. Physical dependence can set in within a few weeks of regular use, but it is all the less likely the shorter and more irregular the use (Brunner et al., 2025).
What that implies. The prescriber can often stop more simply, and information about what may happen is usually enough. Occasional use that serves to avoid situations falls under the manual for the anxiety disorder concerned.
2. Physical dependence at the prescribed dose
What you observe. Daily use for months or years, at the prescribed dose, without escalation, with no source other than the prescription. Symptoms appear when the person forgets a dose or tries to cut down, and they conclude from this that they cannot do without it.
What points to it. Use is as prescribed, but it goes on, and stopping produces symptoms. The joint guideline published in 2025 by ten US scientific societies, brought together by the American Society of Addiction Medicine, clearly distinguishes this physical dependence, which is expected and far more common, from use disorder (Brunner et al., 2025).
What that implies. It is the indication of this programme.
3. Use disorder
What you observe. A loss of control: doses that increase without medical advice, prescriptions obtained from several doctors, pills from other sources, use for the effect rather than to sleep or calm down, combinations with alcohol or other substances, and continued use despite clear consequences.
What points to it. The loss of control and use outside the prescription, and not merely the presence of symptoms on stopping.
What that implies. Addiction care, which begins with an assessment of the dependence and of the pattern of use (Haute Autorité de santé, 2015), and which may lead to inpatient withdrawal when the risk cannot be managed on an outpatient basis (Brunner et al., 2025). This programme can be added to it, with the adaptations in section 30.
4. A prescription that remains justified
What you observe. A long-term treatment for which the doctor considers, after review, that the benefit still outweighs the risk: certain epilepsies, end-of-life care, certain severe anxiety conditions that have resisted other treatments (Brunner et al., 2025).
What points to it. An explicit and considered medical decision, and not an automatic renewal of the prescription.
What that implies. The programme does not push for stopping. A partial reduction may be the goal, and it is decided with the prescriber (section 30).
The three other things to have ruled out
An active, untreated anxiety or depressive disorder. Among long-term users followed in general practice, the great majority have a current anxiety or depressive disorder: in a sample of more than four thousand French patients, 85% had at least one current diagnosis, most often a major depressive episode or generalised anxiety disorder (Pélissolo et al., 2007). It should be looked for before talking about stopping.
Insomnia that has another cause. Sleep apnoea, in particular, presents as insomnia or fatigue, and it increases the risks of the medication itself (Brunner et al., 2025); screening for it is described in the section on differential diagnosis of Treating chronic insomnia in adults: a therapist's manual.
Cognitive impairment. In an older person, forgetfulness and slowness may come from the medication, from an incipient neurocognitive disorder, or from both; the answer requires a medical opinion (section 30).
The three sorting questions
"How long have you been taking it, and how: every evening, or only on some days?" It separates short-term or irregular use from prolonged daily use.
"What happens when you forget it or try to take less?" It picks up physical dependence, and it brings out what the person has concluded from their previous attempts.
"Do you ever take more than planned, or get it other than through your prescription?" It picks up use disorder. Ask it without suspicion, as a routine question: it is asked of everyone.
4. What prolonged use produces: tolerance, dependence, withdrawal, rebound, relapse
This section gives the words that allow the person to understand what is going to happen to them. Each of these five notions calls for a different response.
Tolerance
Over time, the benefit of the medication declines, while its risks persist or increase (Brunner et al., 2025). That is why the balance between benefit and risk, favourable over a few weeks, is not established beyond that (Lader, 2011). Many people notice this without putting it into words: the pill no longer makes them sleep as it did at the beginning, but they cannot fall asleep without it.
Physical dependence
This is the body's adaptation to the regular presence of the medication: the receptors on which it acts become fewer or respond differently. It is the expected consequence of regular use, it can set in within a few weeks, and it varies greatly from one person to another (Brunner et al., 2025). It says nothing about the person's character and it is not the same thing as an addiction.
Withdrawal
These are the symptoms that appear when the dose goes down or stops, because the body, adapted to the medication, has to readjust. They can appear even when the medication has been taken at the prescribed doses (World Health Organization, 2022). The most common ones resemble anxiety: nervousness, insomnia, irritability, muscle tension and pain, tremor, sweating, palpitations, headaches, nausea. Others are more disconcerting because they are new: excessive sensitivity to light, noise or smells, pins and needles, a sense of unreality or of being detached from oneself, difficulty concentrating, nightmares (Brunner et al., 2025). The severe forms, namely seizures, confusion and hallucinations, are seen mainly after abrupt discontinuation of high doses; they are a medical emergency (section 9).
Rebound
This is the temporary return of the symptom for which the medication was prescribed, usually defined as more marked than before treatment (Lerner and Klein, 2019): a few worse nights, sharper anxiety for a few days. The guidelines distinguish it from withdrawal and from the lasting return of the disorder (Brunner et al., 2025), and the Haute Autorité de santé (2015) asks that the person be informed of it in the days following the last dose.
Relapse
This is the lasting return of the original disorder, with no link in time to a step: generalised anxiety or insomnia that settles in again and persists for weeks after the end of the taper. It calls for a reassessment and a treatment of the disorder in its own right, not the automatic resumption of the pill (Haute Autorité de santé, 2015).
Why the distinction matters, and why it is difficult
The most common withdrawal symptoms resemble those of the original disorder, and telling them apart is often difficult, including for the doctor (Brunner et al., 2025). A few clues nevertheless point the way: the time of onset in relation to a reduction, the novelty or unusual intensity of the symptom (National Institute for Health and Care Excellence, 2022), and how it evolves when the taper pauses (Brunner et al., 2025). Section 27 turns them into four questions that the person learns to ask themselves on their own. What is at stake is concrete: reading a withdrawal symptom as the return of the disorder leads people to go back to their previous dose (section 5).
Protracted symptoms
Some people describe symptoms that persist for months after stopping, sometimes longer: anxiety, sleep disturbance, sensory or muscular symptoms. Ashton (1991) described them long ago, stressing that they are difficult to define and to measure, and that they do not all reflect withdrawal in the pharmacological sense. The joint guideline recognises that they exist, indicates that they tend to resolve with time and that psychological support can help during this period, and advises against reintroducing the medication to treat them (Brunner et al., 2025). Their frequency and their mechanisms are not established. In practice, you name them without announcing them as likely, you distinguish them from a relapse, and you do not leave a person alone with them (section 30).
5. What keeps the use going
The original reason, a period of insomnia, a bereavement, an anxiety crisis, has often faded long ago. The use continues for other reasons, and it is these that the programme targets.
Immediate relief
Each dose brings relief, and that relief comes quickly: it soothes the evening's anxiety, it spares the apprehension of the night, and, in a person who is physically dependent, it makes the first signs of withdrawal disappear. A behaviour followed by rapid relief is reinforced: this is the mechanism of all habits that bring relief, and it requires no particular weakness.
The reading of symptoms
A missed dose or an attempt to cut down produces withdrawal or rebound symptoms. The person reads them as proof that they need the medication, they go back to the dose, the symptoms disappear, and the conviction grows stronger. The joint guideline notes that many patients make this reading, and insists on clear information, throughout the taper, about what may happen (Brunner et al., 2025). This is the central loop. It is undone by learning to read the symptoms differently (section 27).
The pill as a safety net
The pill kept in the handbag, on the bedside table or in the glove compartment reassures even when it is not taken. Like any safety behaviour, it prevents the person from discovering that they would have coped without it. Panic disorder and agoraphobia: a therapist's manual describes this mechanism in detail in its section devoted to safety behaviours; it applies here as it stands.
The evening gesture
In clinical practice, many people describe the bedtime pill as a ritual: the glass of water, the pack, the gesture, then bed. The gesture seems to have become a signal for bedtime in its own right. This is a clinical observation, which the trials have not measured as such; it explains why session 3 replaces the ritual instead of simply removing it.
Beliefs
"I can't sleep without it." "Without my treatment, I'll relapse." "I'm a nervous person, I'll always need it." "If I stop, it will be terrible." These beliefs have often been confirmed by an attempt to stop abruptly, gone through alone, which produced exactly what they predicted. They are worked on by putting them to the test during the taper, and not by arguing with them (sessions 2, 6 and 8).
Fear of withdrawal
Many people put off stopping because they dread the symptoms or the return of their disorder (Brunner et al., 2025). This fear is legitimate, and it keeps the use going on its own: not trying protects the person from what they dread. It is dealt with by informing, by preparing, and by giving the person control over the pace, which the Haute Autorité de santé (2015) also asks for.
The renewal of the prescription
The prescription is often renewed without the question of stopping being raised, whereas the Haute Autorité de santé (2015) asks for that question to be raised at every renewal. The programme replaces this automatic renewal with an explicit decision.
The complete circle
A difficult night or a surge of anxiety calls for the pill, which brings relief, and the relief reinforces the use. Tolerance reduces the effect, but dependence sets in. A missed dose produces symptoms, read as proof of need; the belief hardens, the fear of stopping grows, and the prescription is renewed.
This is the diagram you draw at session 2.
Frequently asked questions
The person asks me how much to cut, and when. What should I say?
That this is decided with their doctor, and that your role is to help them prepare that conversation, keep to the plan and get through the steps. Help them write down their questions for the doctor, and give no opinion on doses, not even an approximate one (section 12).
The person has already stopped all at once several times, without any problem. Is a programme really needed?
If the use was short-term or irregular, perhaps not: information and the doctor's opinion are often enough (section 3). If the use has been daily for a long time, the absence of problems in the past guarantees nothing, and abrupt discontinuation is still advised against (section 9).
How can I tell whether it is withdrawal or the return of their anxiety?
You do not always know, but four questions point the way: the timing in relation to the last step, the novelty or intensity of the symptom, how it evolves at a constant dose, and the emergency signs. Section 27.
The person took a pill again after stopping. Does everything have to start again?
No: it is a slip, not a relapse. Look with them at the situation that preceded it, return to the approach written on worksheet 8, and see them sooner than planned if it happens again (section 29).
The doctor does not want to taper. What should I do?
Respect the decision, which may be well founded (section 3), and pass on to the doctor what the person wants and what you observe. A decision shared between the person and their doctor remains the condition of the programme; you do not taper without the doctor.
The person is 85 and has been taking their sleeping pill for thirty years. Is it really sensible to stop it?
It is in this population that stopping is most worthwhile, because of falls, and that the trials obtained the best results; the joint guideline proposes tapering in most people aged 65 and over (Brunner et al., 2025). The taper is simply slower, and those close to them more present. Section 30.
Do these medications cause Alzheimer's disease?
The link is debated: one study found an association, another does not support a causal relationship (Billioti de Gage et al., 2014; Gray et al., 2016). Say so in those terms, without using it to frighten. The risk of falls and the risk of road traffic accidents, on the other hand, are well documented (section 9).
Sleep is not improving, a month after stopping. Is this a failure?
No: the improvement may take several months (Morin et al., 2004). Keep up the sleep rules, repeat the ISI at the boosters, and if the insomnia persists at six months, treat it in its own right with Treating chronic insomnia in adults: a therapist's manual.
Is therapy really useful, since tapering alone also works?
Tapering is the mainstay, and it often works on its own (Oude Voshaar et al., 2003). Therapy adds a clear advantage in the short term (Darker et al., 2015), including in people who take a sleeping pill for insomnia (Takaesu et al., 2019), but this advantage fades beyond a few months (section 11). This programme is therefore intended first and foremost for people in whom tapering alone has failed or is at risk of failing.