What types of exposure therapy are there? Five modalities

What types of exposure therapy are there?

The question has a short answer and a useful one. The short answer is five. The cognitive-behavioural literature describes five types of exposure therapy in routine clinical use, and all five rest on the same underlying principle. The useful answer arrives later, when someone has to decide which of them the person sitting opposite actually needs.

• In vivo: the patient confronts the real stimulus in its own setting, with the clinician present or through homework agreed between sessions.
• Imaginal: the scene is rebuilt through a script and narration, for stimuli that admit no physical presence.
• Interoceptive: the stimulus is the bodily sensation itself, brought on by brief, rule-governed exercises.
• Virtual reality: the scene is simulated in a headset and the clinician governs its variables from a console.
• Prolonged: a structured trauma protocol that combines imaginal revisiting of the memory with a list of in vivo tasks.

That list is what any search engine returns, and it carries a problem: it mixes two planes. Four entries name a route of administration — where the stimulus sits and how it reaches the patient — while the fifth names an entire protocol, with its own sequence, session count and components. Sorting the types of exposure therapy along that axis is what turns a list of names into a basis for choosing.

The practical difference shows up in a single example. A commercial flight cannot be repeated three times in an afternoon and it does not pause halfway through take-off, so the in vivo route leaves almost nothing to grade; the virtual reality flight phobia scenario, by contrast, replays that take-off as often as the plan calls for. The fear is the same. What changes is what the clinician can do with it.

Route of administration versus therapeutic mechanism

All five types of exposure therapy work on the same learning. The inhibitory learning framework set out by Craske and colleagues (2014), and updated in 2022 towards inhibitory retrieval, holds that extinction does not erase the original fear association: it builds a competing association that has to be retrievable at the moment it matters. What is therapeutic, then, is not how long the patient endures, but the gap between what they predict and what actually happens.

This is why expectancy violation is the design criterion for the session. Clinician and patient agree on a concrete, testable prediction — how long the sensation will last, what the dog will do, whether the carriage can be left — and the scene is built so that the prediction is disconfirmed. That operation can be assembled through any of the five routes, which is the deeper reason this taxonomy does not rank therapeutic quality.

Habituation deserves a precise placing. It was the dominant explanatory model for decades, in the line of the emotional processing account of Foa and Kozak (1986), and it remains a phenomenon visible inside the session: arousal rises and then falls. What later research qualified is its predictive value, because a fall in subjective units of distress within the session forecasts long-term outcome poorly. A patient can leave the room calm and come back the following week just as avoidant.

Translated into the taxonomy: the types of exposure therapy differ in their route of administration, not in the active ingredient. What each contributes is a different degree of access to the stimulus and a different margin of control over it, and that is exactly what the indication turns on. The logic of graded exposure, with its hierarchy and its record-keeping, applies identically across all five.

A decision matrix between types of exposure therapy: four criteria

The choice among types of exposure therapy reduces to four questions, and they are best asked in this order, before any catalogue of materials is opened.

• Stimulus accessibility: is it available when the plan needs it? A lift, a placid dog or a crowded square can be arranged; turbulence, a thunderstorm or an auditorium holding two hundred people cannot.
• Parameter control: which variable does this case need graded with precision, and who governs it? Distance to the animal, height, the number of eyes on the patient, minutes of enclosure.
• Transfer to the natural environment: how much work outside the consulting room will be needed afterwards for the learning to hold where the problem actually occurs?
• Logistical cost and tolerance: travel, third parties involved, diary space, and what the patient is willing to face in the first week.

The four answers do not always point to the same route, and that is where clinical craft comes in. Accessibility tends to favour simulation or imagination; transfer to the natural environment eventually demands the real setting. When the two collide, the parameter decides: if the variable this case needs graded cannot be governed through a given route, that route is unusable, however convenient it looks.

Patient tolerance enters the decision matrix as a clinical criterion, not as a concession. An indication the patient will not accept is an indication that never gets carried out, and the route accepted today opens the one refused a month ago. Before proposing the simulated route, the clinician also reviews the documented contraindications for virtual reality exposure, with the same care they would apply to the risk of travelling out to a real setting.

The simulated route: virtual reality as the fourth entry

The fourth entry on the map of types of exposure therapy adds no new mechanism: it adds control. In a simulated environment the clinician fixes the height, the distance, the number of figures looking at the patient or the density of the traffic, then repeats the same scene with a single variable moved. That is what the in vivo route concedes only rarely, and what the imaginal route leaves to the patient's capacity for visualisation.

The second contribution is access. Stimuli that cannot be booked for a given hour — a take-off, a storm, an operating theatre, a full auditorium — come into the consulting room without depending on the weather or on anyone else's permission. Logistical cost shifts from travel to equipment, which is a predictable cost and one that can be written down over several years.

On efficacy, the honest reading is equivalence. The meta-analysis by Powers and Emmelkamp (2008) placed the simulated route on a par with in vivo exposure across the anxiety disorders, and Carl and colleagues (2019) found no significant difference between the two: equivalence, not advantage. The narrative review by Maples-Keller and colleagues, published in Harvard Review of Psychiatry in 2017, surveys the whole field — specific phobia, social anxiety, panic, trauma — and lands in the same place: the headset widens access to the stimulus, it does not change the therapy. The figures themselves sit in the meta-analytic comparison against in vivo exposure.

The argument should not be inverted. Among the types of exposure therapy, the simulated route does not depose the in vivo route: it sits at the same level of efficacy with a different profile of control and cost. When the real stimulus is five minutes away and the critical parameter can be governed there, a catalogue of scenarios is a luxury the case does not need.

Prolonged exposure: a protocol before it is a route

The fifth entry among the types of exposure therapy belongs to a different logical level. Prolonged exposure is a closed protocol for post-traumatic stress, with a defined session count, sequence and set of components, combining imaginal revisiting of the memory with a list of in vivo tasks aimed at what the patient avoids. It does not compete with the other four entries: it draws on them.

Its theoretical root is the emotional processing account of Foa and Kozak, and its institutional backing comes from the ISTSS (2018) recommendations, which place it among the first-line interventions for trauma. Those same recommendations are explicit about the frame: they describe a protocol that requires specific training and a clinician who owns the case, not a technique to be improvised inside a supportive session.

The practical consequence is that two distinct decisions are in play, and much of the confusion about types of exposure therapy comes from fusing them. One is the route — where the stimulus sits and who controls its variables. The other is the protocol that orders the whole sequence of treatment. Fusing them produces the familiar error of announcing prolonged exposure when what took place was twenty minutes of imaginal work. The version of the protocol supported by virtual reality in prolonged exposure for PTSD carries a literature and a set of cautions of its own.

Combining modalities and transfer to the natural environment

Real protocols rarely use only one of the types of exposure therapy. In panic disorder the interoceptive and in vivo routes are chained: the feared sensation is first brought on through brief exercises, and the places avoided for fear of that sensation are confronted afterwards. NICE CG113 frames these interventions as components of a cognitive-behavioural treatment, never as free-standing techniques.

The most common sequence in specific phobia starts on the route with most control and ends in the natural environment [CITATION TO VERIFY]. The reason is mechanistic: among the optimisation strategies Craske describes are variability of the scene, work across multiple contexts and retrieval cues, and all three point towards having the inhibitory learning retrievable where the patient lives. All three apply to the five types of exposure therapy, and no simulation exempts anyone from the final stretch.

Transfer to the natural environment is, in fact, the point at which the choice of route gets judged. A case that progresses in the headset and never moves outside the consulting room does not have a technology problem: it has an incomplete plan. The between-session task record is what converts the chosen route into a clinical result.

The beginner's error is not picking the wrong route: it is picking by what happens to be in the room. Whoever has just bought equipment tends to take every case into simulation, and whoever has none resolves by imagination cases that called for physical presence. The mistakes clinical supervision picks up most often repeat that pattern with uncomfortable regularity. Öst (1989) also showed that in specific phobia a single long session can be enough, a reminder that fine-grained grading is a tool rather than a doctrine.

Recording the decision and revisiting the criterion

The choice among types of exposure therapy gets written down. A useful note holds four items: the route chosen, the critical parameter to be graded, the prediction being tested, and the discarded alternative with the reason for discarding it. With that, any colleague opening the clinical record understands the plan without having to reconstruct it from memory.

The criterion is revisited like the rest of the plan. If the critical parameter stops being governable — because the patient already tolerates the highest scene in the catalogue, or because the real stimulus is no longer available — the route changes. The types of exposure therapy are not schools one signs up to: they are resources that rotate within a single case, and the rotation is part of the treatment.

What does not rotate is who decides. The taxonomy orders the available material; the indication requires individual assessment, a diagnosis made by the clinician, and informed consent that explains the route in terms the patient understands. A route being available does not make it indicated, and that logical leap is the one to watch for whenever the choice is made from the equipment rather than from the case.

To see how a hierarchy organised by parameters is configured inside a working system, and what each session leaves on the record, it is possible to book a demonstration of the clinical environment. The decision about route stays with the professional; the software carries it out and documents it.

This article is for informational purposes for psychology professionals. It is not clinical advice for any individual case and does not replace the judgment of the licensed psychologist in charge. VRET is professional clinical-support software, not a CE-marked medical device.