Flooding · Atlas

1. Introduction

Version followed: symptom-focused flooding as practised by Isaac Marks and colleagues at the Maudsley/Institute of Psychiatry in the 1970s. This means prolonged, continuous, high-intensity exposure to the client's own feared cues (in vivo by preference, imaginal as a variant), with escape and avoidance blocked and the session continued until fear falls. Modern consent, safety and inhibitory-learning refinements are added and marked [synthesis] or cited separately. Stampfl and Levis's implosive therapy, which builds exaggerated scenes from psychodynamic cues the therapist hypothesises, is a sibling method and is not covered as a technique here. [source: Emmelkamp 2005 SAGE encyclopedia entry; Henry 1975 thesis; Marks, Boulougouris & Marset 1971]

Flooding confronts the client with a strong dose of feared cues at or near the top of the hierarchy, rather than working up step by step, and keeps them there with no escape. The original learning-theory account was Mowrer's two-factor model: fear is conditioned to cues and kept alive by avoidance. Blocking avoidance while the cues are present lets the fear response extinguish. By the early 1970s Marks's group had reframed this as habituation: the client gets used to the stimulus, and high anxiety may be a by-product rather than the active ingredient, so prolonged exposure without escape is the essential element. The current mainstream account is inhibitory learning: the client learns new safety information that competes with the old fear memory, which is helped by disconfirming the specific outcome they fear. [source: Emmelkamp 2005; Henry 1975 citing Boulougouris et al. 1971, Marks 1972–73; Craske et al. 2014 via Good Medicine summary]

Historical lineage: Malleson (1959, Lancet) told a student with examination panic to stop fleeing the dread and face it fully. Stampfl developed implosion from 1957 and published it with Levis in 1967. Wolpe is credited (Wikipedia, citation flagged) with an in vivo case of a car-phobic girl driven around for hours. Marks's group then tested and refined flooding against desensitisation. In current practice it survives mostly inside intensive or massed exposure formats, not as a stand-alone named therapy. [source: Henry 1975; EBSCO Research Starter; Wikipedia] [uncertain: Malleson's original paper is not drawn on. His instructions are known here only through Henry 1975.]

2. Framework

Flooding is a technique delivered inside a behavioural treatment, not a manualised multi-phase programme. No single official manual exists. The phases below are assembled from the Maudsley trials and later exposure practice. [source: Emmelkamp 2005; Marks et al. 1971, 1975; Yuksel et al. 1984] [synthesis for the phase structure]

Phase A – Assessment and suitability (1–2 sessions, 50–60 min).

Goals: map feared cues, avoidance, escape and safety behaviours, rituals (if OCD), and the feared outcome. Screen medical and psychiatric contraindications. Decide whether flooding is the right pace for this client or whether graded or rapid-graded exposure fits better.
Move on when there is a clear target stimulus set, no unresolved contraindication, and the client understands the alternatives. [synthesis]

Phase B – Preparation and consent (1 session).

Goals: give the rationale, set up a rating scale (Marks's group used 0–8; Wolpe's SUDS is 0–100), choose the starting item (top or near-top), agree the stop rules and how long a session may last, and plan logistics (for in vivo: location, time, transport, a buffer for long sessions).
Move on when there is documented informed consent that names the expected distress and the client's right to stop, plus an agreed plan. [source: Wikipedia, on agreeing the tolerated anxiety level in advance] [synthesis for documentation]

Phase C – Flooding sessions (typically a few to about 10 sessions for phobias; more for OCD).

Format: individual and long. Analogue studies used 20–60 min, but clinical studies that worked used sessions of up to several hours. Stern & Marks (1973) found 80 minutes of continuous exposure beat the same time broken into 10-min blocks. Watson, Gaind & Marks (1972) used about 2 hours of taped flooding. Marks, Hodgson & Rachman (1975) treated chronic OCD with a mean of 23 in vivo exposure sessions, including 4–12 inpatient weeks, then home-based sessions. Sessions are massed, i.e. close together (daily or several a week). [source: Emmelkamp 2005; Henry 1975; Marks et al. 1975] [uncertain: no source cited gives a fixed frequency. 'Daily or several per week' is inferred from intensive and inpatient formats.]
Within-session rule: do not end while fear is at or near peak. Continue until there is a clear, sustained fall in rated fear (classic criterion), or until the client's specific feared outcome has been tested and has not happened (modern criterion). [source: Emmelkamp 2005; Craske et al. 2014]
Move from item to item, or from imaginal to in vivo, when fear at the start of exposure to the current item is low or moderate and the client can stay without safety behaviours. In Yuksel et al. (1984), therapist-aided exposure moved to the next item after a set drop (2 points in the 'rapid' arm, 4 in the 'slow' arm, on 0–8). Rapid progression shortened treatment (mean 4.7 vs 5.8 sessions) without raising peak anxiety or worsening outcome.

Phase D – Self-exposure and generalisation (overlaps Phase C).

Goals: daily self-directed exposure between sessions, in varied settings, with safety behaviours removed. Marks's programmes always paired therapist-aided sessions with self-exposure homework. [source: Yuksel et al. 1984; Marks et al. 1975]
Move on when the client enters the top-item situations alone with only mild fear.

Phase E – Ending and follow-up.

End-of-treatment criterion in Yuksel et al.: tolerating the top hierarchy item with minimal discomfort, or reaching a cap of 8 sessions. Follow-up at about 6 months. [source: Yuksel et al. 1984]
Relapse-prevention plan for return of fear. [synthesis]

3. Guiding principles

Prolonged beats brief. Sessions must be long enough for fear to fall or for the feared prediction to be disconfirmed. Book long slots, and never schedule flooding into a 50-min hour. [source: Emmelkamp 2005; Henry 1975 citing Rachman 1966, Kotila 1969, Stern & Marks 1973]
Continuous beats interrupted. Breaks and early exits act as escape and may sensitise. Keep the stimulus present without interruption once you start. [source: Emmelkamp 2005; Henry 1975]
No escape, no avoidance, but always voluntary. Blocking avoidance is the mechanism. Consent to stay must be renewed by the client, never imposed by force. Agree stop rules beforehand, and the client keeps the right to stop. [source: Emmelkamp 2005; Wikipedia] [synthesis: the voluntariness framing reflects current ethics; early sources do not spell it out]
Start high. Unlike desensitisation, flooding begins at or near the top of the hierarchy. In practice, start at the highest level the client consents to. [source: Emmelkamp 2005; Marks 1974 letter]
Encourage tolerance, don't soothe it away. Marks reported that patients were "vigorously encouraged" to tolerate more anxiety than in desensitisation. Coach the client to stay. Do not reassure them that the feared outcome cannot happen. [source: Marks, BJP correspondence 1974; Wikipedia]
In vivo over imaginal where feasible. Real-life exposure generally outperformed fantasy exposure. Use imaginal flooding when the cue cannot be produced (catastrophes, intrusive images) or as a bridge to in vivo work. [source: Henry 1975 citing Stern & Marks 1973; Emmelkamp & Wessels 1975 abstract]
Relevance matters less than staying. Marks's group found relevant and irrelevant fear scenes worked similarly. Exposure without escape mattered more than evoking maximal anxiety. Do not escalate horror for its own sake. [source: Henry 1975 citing Watson & Marks 1971; Boulougouris et al. 1971]
Too fast can backfire. 'Too rapid exposure' with excessive anxiety seemed to slow progress (Watson, Gaind & Marks). Pace within what the client can stay with. [source: Henry 1975]
Self-exposure carries the gains. Therapist-aided flooding is the start, and daily self-exposure is where most practice happens. [source: Yuksel et al. 1984; Marks et al. 1975]
Relaxation is not the active ingredient. Muscular relaxation added nothing in Marks's OCD trial. Do not teach relaxation as a way to escape during flooding. [source: Marks et al. 1975] [uncertain: Wikipedia describes relaxation as part of flooding; the Maudsley trials did not support that]
Disconfirm the prediction (modern refinement). Elicit what the client expects to happen and design exposure to test it, rather than relying only on fear dropping. [source: Craske et al. 2014 via secondary summary]
Measure every session. Rate fear at regular intervals so both you and the client can see the curve. [source: Yuksel et al. 1984; Wolpe SUDS convention] [synthesis on frequency]

4. Intervention inventory

The primary sources describe flooding as a small set of procedures (in vivo flooding, imaginal flooding, taped flooding, response prevention, modelling, self-exposure) rather than a long list of named exercises. Everything below is anchored to those sources. Preparatory and safety steps that the old trials took for granted are spelled out and marked [synthesis]. All example language is illustrative unless quoted. [source: Emmelkamp 2005; Henry 1975; Marks et al. 1971, 1975; Yuksel et al. 1984]

4.1 Functional analysis of the fear

Purpose / when: First sessions. Identify exactly which cues to flood with, and which escape, avoidance and safety behaviours must be blocked.
Steps: (1) List the situations, objects, images, thoughts and bodily sensations that trigger fear. (2) For each, record what the client does to escape, avoid or neutralise it: leaving, distraction, holding on, a companion, medication carried 'just in case', checking, washing. (3) Elicit the feared outcome in the client's own words ('I'll faint', 'I'll lose control', 'the house will burn'). (4) Rank items 0–8 or 0–100 for expected fear to build a hierarchy, and mark the top 2–3 items as flooding candidates. (5) Note practical access: can the cue be produced in real life, in a session-length block?
Example language (illustrative): "Walk me through the last time you were near a lift. What did you notice first, what did you expect would happen, and what did you do to get through it or get out?"
Watch for: Hidden or subtle safety behaviours (sitting near exits, mental counting). If you don't find them now, they will quietly undermine sessions. A vague feared outcome makes prediction-testing hard later. [source: Marks et al. 1975 on rituals; Craske et al. 2014 on safety signals] [synthesis for step detail]

4.2 Suitability and contraindication screen

Purpose / when: Before offering flooding over graded exposure. Flooding is the most intense exposure format, and some clients should not start with it.
Steps: (1) Medical: ask about cardiovascular disease, recent cardiac events, uncontrolled asthma, seizure disorders, pregnancy and other conditions where sustained high arousal or exertion is a risk. Get GP or physician clearance where relevant. Henry (1975) excluded a participant with a heart murmur. (2) Psychiatric: active psychosis, current suicidal intent or recent serious self-harm, current substance intoxication or dependence, severe dissociation, acute life crisis. Each calls for stabilisation first or a gentler format. Assess suicide risk with your usual structured clinical interview. (3) Capacity and voluntariness: the client must be able to consent and must choose this freely. Do not use flooding for mandated or coerced clients, or with children unless there is guardian consent, the child's own assent, and a clear reason for not using a graded or one-session format. (4) Preference: offer graded or rapid-graded exposure as an equally legitimate option. (5) Record the decision.
Example language (illustrative): "There are two ways we can do this: step by step, or starting near the top and staying until the fear comes down. Both work. The second is faster but much more uncomfortable. Before we choose, I need to check a few health questions."
Watch for: A client who agrees out of shame or to please you. Look for real choice, not compliance. [source: Henry 1975 on cardiac exclusion; EBSCO on risk of worsening] [synthesis for the screen list] [uncertain: no primary source cited gives a formal contraindication list]

4.3 Rationale and informed consent

Purpose / when: Phase B. The client needs a model that makes staying in the situation make sense, plus an honest picture of the distress involved.
Steps: (1) Explain the avoidance cycle: escaping brings short-term relief, but it stops the client from learning that the fear passes and the catastrophe does not happen. (2) Explain the flooding logic: staying with the cue, without escape, long enough for fear to fall and for predictions to be tested. (3) Describe the session honestly: a long session, fear likely to peak early and high, then come down. Tell the client which sessions run 1–3 hours. (4) Name the risks: intense distress, temporary worsening, possible return of fear later, and possible dropout. (5) State the right to stop and how stopping will be handled (see 4.4). (6) Get consent in writing or in the notes. Offer time to think.
Example language (illustrative): "Each time you leave a frightening place, your brain files it as 'that was dangerous and I escaped just in time.' Flooding gives it the other experience: you stay, the fear rises, and then it falls without anything terrible happening."
Example language (illustrative): "Marks's patients described it as getting used to it. That's what we're aiming for, and it usually takes longer than people expect."
Watch for: The client hearing 'you'll be forced to stay'. Correct this immediately. Signs of understanding: the client can explain in their own words why leaving early would undo the work. [source: Henry 1975 citing Marks 1973 on 'getting used to'; Emmelkamp 2005; Wikipedia] [synthesis for script]

4.4 Stop rules and tolerance agreement

Purpose / when: Before the first flooding session. This reconciles the no-escape principle with voluntariness.
Steps: (1) Agree in advance how high a fear level the client will tolerate, and that the session aims to continue until fear falls. (2) Agree a clear stop signal (word or gesture) for a genuine wish to end. Explain that if they use it, you will first ask for a short extension (e.g. a few more minutes) or a step down to a slightly easier version of the cue, so the session ends with fear coming down rather than mid-escape. The client's final 'stop' is always honoured. (3) Agree medical stop criteria, such as chest pain or fainting. (4) Agree what the therapist will and won't do: coach and encourage, not reassure that nothing bad can happen, not physically restrain or block exits.
Example language (illustrative): "If you say 'stop', I'll ask you to stay a little longer, or we'll step back to something a bit easier and stay there until it settles. But if you say stop a second time, we stop. You are always in charge of that."
Watch for: Therapists being tempted to override a repeated stop. Never do it. An agreement that rules out stopping is not valid consent. [source: Wikipedia on pre-agreed tolerated level] [synthesis for the step-down and the stop signal]

4.5 Fear ratings and in-session monitoring

Purpose / when: Every flooding session. Tracks the fear curve, decides when to end or advance, and gives the client evidence of change.
Steps: (1) Use one scale throughout: 0–8 (Marks/Yuksel) or 0–100 SUDS. (2) Take a rating at the start, then every 5–10 minutes, and at peak and end. (3) Write down the numbers, and optionally graph them for the client afterwards. (4) If you use the expectancy approach, also rate 'how likely is [feared outcome] right now?' (0–100%) at the start and end.
Example language (illustrative): "Quick number, 0 to 8: how frightened right now? ... And how sure are you, 0 to 100, that you will faint if you stay?"
Watch for: Ratings flat at zero while the client looks tense. This suggests mental avoidance or minimising. Ratings that never fall over a long session suggest hidden safety behaviour (see Section 5). [source: Yuksel et al. 1984; Craske et al. 2014] [synthesis for interval]

4.6 Eliciting the feared prediction (modern refinement)

Purpose / when: Before each exposure. Turns the session into a test the client can see the result of, and supports inhibitory learning.
Steps: (1) Ask what the client expects will happen if they stay: specific, observable, time-bound. (2) Rate how strongly they believe it. (3) Design the exposure so that the outcome could in principle occur, so the test is fair: e.g. stay long enough, no safety behaviours. (4) Afterwards, ask what actually happened and what they learned. Hold back from summarising it for them.
Example language (illustrative): "What's the worst you think will happen if we stay in the supermarket queue for the full hour? How would we know if it happened?"
Watch for: Predictions that drift ('I'd have fainted if we'd stayed longer'). Design the next session to test the new version. [source: Craske et al. 2014 via Good Medicine summary] [uncertain: this is a post-flooding-era refinement, not part of classic flooding]

4.7 In vivo flooding session (core procedure)

Purpose / when: Phase C. This is the main intervention for phobias, agoraphobic avoidance and OCD triggers that can be produced in real life.
Steps: (1) Book a long, open-ended slot (90 min to several hours) and agree in advance that the session continues until fear falls. (2) Briefly review the rationale, stop rules and the feared prediction. (3) Take a baseline rating. (4) Bring the client into contact with the agreed high-level cue (top or near-top item): e.g. enter the crowded shopping centre, hold the contaminated object, sit in the stationary then moving lift. (5) Keep contact continuous. No leaving, no distraction, no safety items, no rituals. (6) Coach to stay and to attend to the situation, not to escape mentally. Give minimal reassurance about safety. (7) Rate every 5–10 minutes. (8) Expect a rise, a peak and then a decline. Continue past the peak until fear has clearly and steadily fallen (e.g. roughly half of peak, or low and stable for several readings) and/or the feared prediction has been tested. (9) If time allows, raise intensity within the session (closer, longer, alone). (10) End only on a falling or low fear rating, never at peak. (11) Debrief (4.18) and set self-exposure (4.17).
Example language (illustrative): "Stay with it. Let the fear be as big as it wants to be. You don't need to fight it or get away from it."
Example language (illustrative): "You're at 7. Notice that, and stay put. We'll check again in five minutes."
Example language (illustrative): "It's come down to 3 and you're still here, without holding the rail. What does that tell you?"
Watch for: Early exit (the biggest risk; see Section 5). Fear that won't fall points to hidden safety behaviour, rumination, dissociation or a session that is too short. Faintness: blood-injury phobia has a vasovagal pattern, so use applied tension rather than plain flooding (that comes from Öst's applied-tension work, not this guide's sources) [uncertain]. Signs it is working: within-session fall, a lower peak at the next session, and the client starting top items more willingly. [source: Emmelkamp 2005; Henry 1975 citing Stern & Marks 1973; Marks 1974 letter; Wikipedia] [synthesis for step detail and endpoint numbers]

4.8 Response prevention (for compulsions)

Purpose / when: OCD, or any presentation where a ritual or neutralising act ends exposure early. It was used alongside in vivo exposure in the Maudsley OCD work.
Steps: (1) List every ritual and neutraliser: washing, checking, reassurance-seeking, mental reviewing, counting. (2) Agree which are suspended, for how long, and with which exceptions (e.g. normal hygiene routines). (3) During and after the flooding exposure, the client refrains from them. In Marks et al. (1975) this was self-imposed, not enforced by staff. (4) Where useful, involve family members in not giving reassurance or doing rituals on the client's behalf, with the client's agreement. (5) Review lapses without blame and re-commit.
Example language (illustrative): "After you touch the bin, the deal is no handwashing until our next check-in. If the urge feels unbearable, ring me rather than wash, and we'll work out the next step together."
Watch for: Covert or mental rituals replacing overt ones. Family accommodation. Depression emerging during treatment (common in Marks et al. 1975). Assess and treat it. Signs it is working: urges fade faster after each exposure, and the client needs fewer reassurance requests. [source: Marks, Hodgson & Rachman 1975 abstract] [synthesis for steps and phone contact]

4.9 Therapist modelling

Purpose / when: Optional, at the start of an in vivo session, when the client cannot begin contact without seeing it done.
Steps: (1) Do the feared action calmly yourself first: touch the 'contaminated' surface, hold the spider jar, step into the lift. (2) Invite the client to copy immediately. (3) Withdraw your modelling and physical presence as soon as possible so the client's learning does not depend on you.
Example language (illustrative): "Watch me put both hands on the door handle and leave them there. Now you."
Watch for: The therapist becoming a safety signal. Modelling gave no group-level advantage over exposure alone in Marks et al. (1975), though it may help some individuals. Rachman, Hodgson & Marks (1971) found modelling and flooding comparably effective. [source: Marks et al. 1975; Henry 1975 citing Rachman et al. 1971]

4.10 Imaginal flooding (therapist-narrated)

Purpose / when: When the cue can't be produced in real life or safely (catastrophic outcomes, rare events, obsessional images), or as preparation before in vivo flooding. Most early clinical flooding was imaginal.
Steps: (1) Using the functional analysis, write a scene that includes the client's real cues and feared consequences, in first person and present tense, with sensory detail (sights, sounds, body sensations) and the thoughts the client fears. (2) Have the client close their eyes. Narrate the scene continuously and vividly at a high fear level, without escape within the scene. (3) Ask for ratings every few minutes without breaking the image. (4) Repeat or extend the scene. Hold the most distressing points. Keep going, uninterrupted, until ratings clearly fall. Clinical sessions ran much longer than the 20–60 min of analogue studies. (5) Debrief and ask what the client noticed. (6) Move to in vivo exposure as soon as feasible.
Example language (illustrative): "You're in the lecture hall. The paper is face down. You turn it over and the first question means nothing to you. Your heart is pounding. You think: I'm going to fail, everyone will know. Stay right there. What's your number?"
Watch for: Clients 'switching off' the image (ask what they are seeing). Scenes that escalate into gratuitous horror: Watson & Marks found irrelevant fear scenes as effective as relevant ones, so the shock value isn't what drives change. Imaginal work was generally less effective than in vivo. Trauma memories: for PTSD use a dedicated protocol (e.g. prolonged exposure) rather than generic flooding [synthesis]. Signs it is working: ratings fall within and across sessions, and the scene starts to feel 'boring'. [source: Emmelkamp 2005; Henry 1975 citing Watson & Marks 1971, Stern & Marks 1973]

4.11 Recorded imaginal flooding

Purpose / when: To deliver long imaginal sessions, or to extend practice between sessions.
Steps: (1) Record the narrated scene in a session (the client's voice or yours). (2) The client listens in a private, undisturbed setting for the agreed full duration (Watson, Gaind & Marks used about 2 hours), rating fear at intervals. (3) Review the ratings at the next session.
Watch for: Mixed evidence. Some studies found taped flooding ineffective. Clients may stop early or listen while distracted. Use it as an adjunct, not a replacement for live work. [source: Henry 1975 citing Watson, Gaind & Marks 1972; Dee 1970; Fazio 1970] [uncertain]
Example language (illustrative): "Play the whole recording through in one go, with no pausing. Every ten minutes or so jot down your number on the sheet."

4.12 Imaginal-then-in-vivo sequencing

Purpose / when: For agoraphobic or highly avoidant clients who refuse in vivo flooding at first.
Steps: (1) Run one or more imaginal flooding sessions on the target situation. (2) Move the same session, or the next one, into real-life exposure to that situation.
Watch for: Staying in imagination too long because it feels safer. Mathews & Gelder suggested some agoraphobic patients improve in vivo only after fantasy exposure. Emmelkamp & Wessels compared in vivo, imaginal and combined flooding. [source: Mathews & Gelder, BJP correspondence 1974; Emmelkamp & Wessels 1975 abstract] [uncertain: the comparative results are not drawn on in full]
Example language (illustrative): "You've just stayed with the train scene in imagination until it came down. Shall we take that straight to the station now?"

4.13 Rapid-progression massed exposure (graded variant)

Purpose / when: For clients who decline full flooding, or when you want massed exposure with less abrupt entry.
Steps: (1) Build an 8-item hierarchy. (2) In a long therapist-aided session, start at the lowest item. (3) Move to the next item as soon as fear on the current one has fallen by an agreed amount. Yuksel et al. used 2 points on 0–8 in the 'rapid' arm. (4) Continue up the hierarchy within and across sessions. (5) Pair with daily self-exposure. (6) Finish when the client tolerates the top item with minimal discomfort.
Example language (illustrative): "You've dropped from 6 to 4 standing at the escalator. That's our signal: up we go."
Watch for: Rapid pacing shortened treatment (4.7 vs 5.8 sessions; 360 vs 484 min) without higher peak anxiety and with equal outcomes at six months. It is a reasonable default when full flooding is not chosen. [source: Yuksel, Marks, Ramm & Ghosh 1984]

4.14 'Face it rather than fight it' instruction (Malleson)

Purpose / when: At the start of, and during, exposure to panic or dread, for clients who are bracing against or fighting the fear.
Steps: (1) Ask the client to stop trying to escape or suppress the dread. (2) Ask them to deliberately let the fear and the feared images come fully. (3) Have them practise this when panic arises between sessions.
Example language (illustrative): "Instead of pushing the panic away, let it come. Picture the worst of it as fully as you can, and let it run its course."
Watch for: The client using 'letting it come' as a ritual meant to make the fear go away. Malleson's student reportedly found he could hardly frighten himself once he stopped fleeing. [source: Henry 1975 summarising Malleson 1959] [uncertain: original paper is not drawn on; step detail is inferred]

4.15 Identifying and dropping safety behaviours

Purpose / when: Throughout Phase C. Safety behaviours are a subtle form of escape that can block learning.
Steps: (1) Using the functional analysis, list the behaviours: holding on, carrying water or medication, a companion, sunglasses, checking pulse, staying near exits, mental distraction. (2) Agree which to drop at this session. Drop them all for a full flood, or progressively if the client prefers. (3) Check during the session ('Anything you're doing to keep yourself safe right now?'). (4) Drop the next ones in later sessions and in homework.
Example language (illustrative): "This time, the water bottle stays in the car and you stand in the middle of the room, not by the door."
Watch for: New safety behaviours replacing old ones. Fear that only stays low while the safety behaviour is present. [source: Craske et al. 2014 'removal of safety signals' via summary] [synthesis for steps]

4.16 Inhibitory-learning enhancements (modern refinements)

Purpose / when: Phases C–D. These make learning from flooding stick and reduce return of fear. They come from later exposure science, not classic flooding.
Steps: (1) Expectancy violation: see 4.6. (2) Deepened extinction: once two cues have been exposed separately, combine them (e.g. crowded train while doing an interoceptive exercise). (3) Variability and multiple contexts: change the location, time, stimulus type and companions across sessions and homework. (4) Affect labelling: ask the client to describe their feelings aloud during exposure. (5) Retrieval cues: for relapse moments, teach the client to recall a successful exposure, or carry a cue that was present during it. (6) Occasional reinforced extinction is described by Craske et al. but is rarely appropriate clinically. Do not plan it without specialist supervision.
Example language (illustrative): "Say out loud what you're feeling right now, in a few words." / "Next week, same exercise, different shopping centre, different time of day."
Watch for: Variability that raises fear so much the client starts escaping again. Bring the intensity down while keeping the variety. [source: Craske et al. 2014 via Good Medicine summary] [uncertain: full paper is not drawn on; step detail is from the summary plus synthesis]

4.17 Self-exposure homework

Purpose / when: From the first flooding session onwards. This is where most repetition happens.
Steps: (1) Set specific daily tasks: the same or similar items to the session, for a long continuous period. Stay until fear falls, no safety behaviours. (2) Give a simple diary: date, task, duration, fear at start, peak and end, feared prediction and outcome. (3) Review the diary first in every session. (4) Increase difficulty and variety each week.
Example language (illustrative): "Every day this week, an hour in town on your own. Don't leave until your number has come down by at least half. Write it down even if it goes badly."
Watch for: Brief entries (5–10 min) that end at peak, which may sensitise. Doing the task with a companion when you agreed 'alone'. A diary filled in from memory. Signs it is working: longer stays, lower peaks, entries the client initiated. [source: Yuksel et al. 1984; Marks et al. 1975] [synthesis for diary fields]

4.18 Post-session debrief

Purpose / when: End of every flooding session. Consolidates learning and checks the client's state before they leave.
Steps: (1) Show the fear curve. (2) Ask what the client predicted, what happened and what they learned. (3) Check current distress, physical state and safety. If the client is still highly aroused or dissociated, stay with them until settled, and screen for self-harm urges if relevant. (4) Plan self-exposure. (5) Make sure the client can get home safely (they should not drive if faint or very shaken).
Example language (illustrative): "Your number peaked at 8 in the first twenty minutes and you finished on 2. What did you learn about what happens when you stay?"
Watch for: 'I only coped because you were there.' Plan solo practice to address this. [synthesis]

4.19 Ending and relapse prevention

Purpose / when: Phase E.
Steps: (1) Confirm the end criterion: top item tolerated with minimal discomfort, alone. (2) Explain that some return of fear is normal after time away, stress or a new context. Wikipedia notes spontaneous return as a known limitation. (3) Write a plan: early signs, an immediate self-exposure response, and 'booster' self-flooding. (4) Book follow-up, e.g. at 6 months as in Yuksel et al.
Example language (illustrative): "If the fear creeps back, treat it as a cue to go straight back in and stay. Avoidance is what will grow it again."
Watch for: Re-emerging avoidance between sessions as the client winds down. [source: Yuksel et al. 1984; Wikipedia] [synthesis]

4.20 Related massed format: one-session treatment (Öst)

Purpose / when: Specific phobias in adults and children, when you want massed, intensive exposure without classic flooding's abrupt top-item entry.
Steps: As described in the published review: a single session of up to 3 hours that combines graduated exposure, participant modelling, cognitive challenges and reinforcement, with the therapist leading the client through the hierarchy as a series of behavioural experiments. A pre-treatment assessment session is used. Train in the protocol before using it.
Example language (illustrative): "What do you think the spider will do if I put the jar on the table between us? Let's find out."
Watch for: This is a distinct, manualised treatment. Use its own manual. [source: Zlomke & Davis 2008 review abstract] [uncertain: the detailed description (Davis, Ollendick & Öst 2009) is not drawn on]

5. Handling problems

The primary flooding literature says little about clinical troubleshooting beyond 'don't interrupt' and 'encourage tolerance'. Most of what follows is marked [synthesis] and draws on general exposure practice.

Client wants to leave at or near peak. Use the pre-agreed stop routine: acknowledge, rate, ask for a short extension, then step down to a slightly easier version of the cue and stay until fear falls, so the session ends on a decline rather than on escape. Honour a repeated stop. Next session, either resume at the step-down level or switch to rapid-graded exposure (4.13). Classic sources warn that ending sessions early can strengthen fear. [source: Emmelkamp 2005; Henry 1975] [synthesis for the step-down]
Fear doesn't fall within the session. Check for covert safety behaviours, mental escape (distraction, dissociation, 'it's not real'), rituals, or a session that is simply too short. Remove these and extend. If you are working imaginally, check the client is actually holding the image. Consider whether the prediction is being tested at all. A session that ends with no fall should still end on a step-down where fear does come down. [source: Henry 1975 on interrupted exposure; Craske et al. 2014] [synthesis]
Fear resets fully between sessions. Space sessions closer together (massed). Add daily self-exposure. Vary contexts. Look for avoidance between sessions that is re-learning the fear. Shift the focus to expectancy violation over within-session habituation. [source: Craske et al. 2014; Yuksel et al. 1984] [synthesis]
Overwhelm, panic beyond what was agreed, dissociation. Ground the client to the present situation while keeping them in contact with the cue: eyes open, describe the surroundings, feel your feet, keep giving ratings. Reduce intensity a step. Don't remove the cue completely unless a medical or safety stop applies. If dissociation recurs, switch to graded exposure with grounding, and reassess for trauma or a dissociative disorder. Watson, Gaind & Marks observed that too-rapid exposure with excessive anxiety seemed to slow progress. [source: Henry 1975] [synthesis for grounding]
Medical event (chest pain, fainting, breathing difficulty). Stop the exposure, give first aid, call emergency services if indicated, and seek medical review before the next session. Faintness in blood-injury phobia calls for applied tension, not flooding. [synthesis] [uncertain: no flooding source cited covers this]
Refusal to start / 'resistance'. Treat it as information, not defiance. Revisit the rationale and the client's goals. Offer the less intense options: imaginal first (4.12), therapist modelling (4.9), rapid-graded exposure (4.13). Never trick, pressure or physically compel. The client choosing a gentler route is a legitimate outcome. [synthesis]
Non-adherence to self-exposure. Check that tasks are specific, scheduled and realistic, and review the diary first. Problem-solve the barrier (time, transport, shame). Shrink the task but keep the duration rule (stay until fear falls). Involve a family member as a reminder (not as a safety companion) if the client wishes. Track with the diary. [source: Yuksel et al. 1984 on mandatory self-exposure] [synthesis]
Dropout risk. Flooding is intensely uncomfortable, and early sources note concern that it may worsen some clients. Reduce risk with honest consent, a clear model, short gaps between the first sessions, and an explicit offer of graded alternatives. If a client stops coming, contact them and offer graded work rather than closing the case. [source: EBSCO Research Starter; Wikipedia] [synthesis]
Rupture: the client feels pushed, betrayed or coerced. Stop and repair before any further exposure. Name what happened, own your part, check whether the stop agreement was honoured, and re-negotiate consent and pacing. Look honestly at whether your own urgency or 'vigorous encouragement' overrode the client's choice. [source: Marks 1974 on encouragement] [synthesis]
Temporary worsening or sensitisation. A 1980 survey of exposure therapists, cited in secondary sources, reported serious negative effects in 9 of 3,493 clients (0.26%). Monitor between sessions. If avoidance or distress clearly increases across two sessions, drop to graded exposure and review the formulation. [source: EBSCO Research Starter citing Shipley & Boudewyns 1980] [uncertain: the original survey is not drawn on]
Depression emerging during treatment. This was common in the Maudsley OCD series. Screen mood regularly, assess suicide risk with standard structured clinical assessment, and treat or refer as indicated (the original series sometimes used medication; prescribing is outside this guide). Pause flooding if mood or risk deteriorates. [source: Marks et al. 1975] [synthesis]
Suicidal ideation or self-harm urges arising. Pause the exposure programme. Do a standard risk assessment and collaborative safety planning, including lethal-means counselling (safe storage or removal), and resume only once the client is stable and consents again. [synthesis]
Family accommodation (OCD, agoraphobia). With the client's consent, bring the family in: explain the rationale, agree that they will stop reassuring and stop doing rituals or tasks on the client's behalf, and give them what to say instead. [source: Marks et al. 1975 on family involvement in home treatment] [synthesis]
Therapist avoidance. Therapists uncomfortable with distress often end sessions early or reassure. Get supervision, rehearse the stop routine, and remind yourself that a falling fear curve is the goal, not a calm session. [synthesis]
Requests to 'flood' someone without their consent (parents, institutions, partners). Refuse. Accounts of nonconsensual flooding include an anecdote attributed to Wolpe of a car-phobic girl driven around for hours (citation flagged on Wikipedia), and 'attachment therapy' practices that Wikipedia says combined nonconsensual flooding with other coercive methods. The APSAC Task Force (Chaffin et al. 2006) rejects treatments that use physical coercion, restraint, domination, deliberately provoked rage or humiliation, and links some attachment therapies to children's deaths. Offer consented graded exposure, one-session treatment, or parent-focused approaches instead. [source: Wikipedia; Chaffin et al. 2006 via NCJRS abstract]
Children and adolescents. Prefer graded exposure or Öst's one-session treatment with the child's assent and parental consent. Full flooding with a child needs a clear justification, assent that can be withdrawn at any time, and supervision. [source: Zlomke & Davis 2008 abstract] [synthesis]

Sources

Sources:

Further reading (not drawn on; known through the sources above) [uncertain]:

Malleson, N. (1959). Panic and phobia: a possible method of treatment. Lancet.
Stampfl, T. G., & Levis, D. J. (1967). Essentials of implosive therapy. Journal of Abnormal Psychology.
Stern, R., & Marks, I. (1973). Brief and prolonged flooding. Archives of General Psychiatry.
Watson, J. P., Gaind, R., & Marks, I. M. (1972). Physiological habituation to continuous phobic stimulation. Behaviour Research and Therapy.
Rachman, S., Hodgson, R., & Marks, I. M. (1971). Treatment of chronic obsessive-compulsive neurosis. Behaviour Research and Therapy.
Emmelkamp, P. M. G., & Wessels, H. (1975). Flooding in imagination vs flooding in vivo. Behaviour Research and Therapy.
Shipley, R. H., & Boudewyns, P. A. (1980). Flooding and implosive therapy: are they harmful? Behavior Therapy.
Craske, M. G., Treanor, M., Conway, C. C., Zbozinek, T., & Vervliet, B. (2014). Maximizing exposure therapy: an inhibitory learning approach. Behaviour Research and Therapy.
Marks, I. M. (1987). Fears, Phobias, and Rituals. Oxford University Press.
Davis, T. E., Ollendick, T. H., & Öst, L.-G. (2009). Intensive treatment of specific phobias in children and adolescents. Cognitive and Behavioral Practice.

← Previous · II-013Bergen 4-day treatment Next · II-015 →Implosive therapy