Aversive Therapy

Abstract

Aversive therapy is a form of behavior therapy that pairs an unwanted behavior, or the cue that provokes it, with an unpleasant stimulus so that the behavior itself comes to evoke discomfort and is thereby suppressed. It applies classical conditioning in the clinic: an appetitive cue such as the taste of alcohol is followed repeatedly by nausea, electric shock, or a vivid aversive image until the cue alone elicits revulsion. Developed in the 1930s and 1940s for alcohol use disorder and later extended to smoking, paraphilias, and self-injurious behavior, it was among the first laboratory-derived behavioral treatments and later among the most contested on ethical grounds. This article surveys its definition, history, conditioning mechanisms, principal techniques, evidence base, and its narrowed contemporary role.

Keywords: aversive therapy, aversion therapy, covert sensitization

What Aversive Therapy Is

Aversive therapy is the deliberate clinical use of punishment and aversive conditioning to weaken a behavior a person wants to stop. Its logic is the mirror image of reward-based treatment: where contingency management strengthens an adaptive behavior by following it with something valued, aversion therapy weakens a maladaptive one by following it, or the cue that triggers it, with something unpleasant. The target is typically an appetitive behavior whose immediate consequences are reinforcing — drinking, smoking, a paraphilic act — so that the disorder persists despite long-term harm. By attaching an aversive outcome to the behavior, the treatment aims to overlay the existing positive association with a competing negative one (#ref-rachman-1969).

Joseph Wolpe placed the method inside the reciprocal-inhibition account of behavior therapy, treating the aversive response as one that could inhibit and displace the appetitive one (#ref-wolpe-1958). Two broad families follow from this premise. In the first, an external aversant — an emetic drug, a brief faradic shock — is physically paired with the cue. In the second, developed by Joseph Cautela, the aversive event is imagined rather than delivered: the client visualizes the unwanted behavior and then a vividly unpleasant consequence, a procedure he named covert sensitization (#ref-cautela-1967).

The approach is easily caricatured, and its history includes genuine abuses, but as a technical matter it is simply the clinical application of the same associative learning that underlies every other conditioning-based treatment. What distinguishes it is the valence of the stimulus it enlists and, consequently, the ethical scrutiny it invites.

Figure 1

The Aversive-Conditioning Contingency

Historical Development

The clinical technique predates a formal learning-theory rationale for it. Isolated nineteenth- and early twentieth-century practitioners had already tried dosing alcoholic drinks with emetics, but the first systematic programme was Walter Voegtlin's conditioned-reflex treatment of alcoholism at the Shadel Sanatorium in Seattle, begun in the 1930s, which paired the sight, smell, and taste of alcohol with drug-induced nausea across a course of sessions (#ref-voegtlin-1940). A decade of follow-up gave the approach something the interpretive therapies of the era lacked: reported abstinence rates from a large treated series, published as an outcome evaluation rather than a case anecdote (#ref-lemere-1950).

The method spread from alcohol to other targets in the 1950s and 1960s. Michael Raymond's report of a fetishism treated by pairing the fetish object with apomorphine-induced nausea became a widely cited demonstration that aversion therapy could be applied beyond addiction (#ref-raymond-1956). Wolpe's reciprocal-inhibition framework gave the growing collection of procedures a theoretical home within behavior therapy (#ref-wolpe-1958), and Cautela's covert sensitization removed the need for drugs or apparatus altogether by moving the aversive event into imagination (#ref-cautela-1967). By the end of the 1960s, Stanley Rachman and John Teasdale could survey the field in the first book-length analysis of aversion therapy and behavior disorders, weighing its mechanisms and its limits (#ref-rachman-1969).

That same period saw aversion therapy applied to homosexuality, a use that the field has since wholly repudiated. Contemporary reviews classify such sexual-orientation change efforts as ineffective and harmful, and the episode is now the standard cautionary case in discussions of the method's ethics (#ref-przeworski-2021).

Conditioning Mechanisms

Aversive therapy rests on classical conditioning. An appetitive cue that already elicits approach — the taste of alcohol, the sight of a cigarette — functions as a conditioned stimulus; the aversant, whether an emetic drug or a shock, is the unconditioned stimulus that elicits discomfort automatically. Repeated pairing transfers the aversive response to the cue, so that the cue alone comes to evoke it. The associative account most often invoked is the Rescorla-Wagner model, in which each pairing increments the cue's aversive strength by an amount proportional to the discrepancy between the aversant's potency and the strength already acquired, producing the decelerating acquisition curve the first demonstration above traces (#ref-elkins-1991).

Two features of the underlying learning shape what the clinic can and cannot do. The first is contiguity: for most conditioning the aversant must follow the cue closely in time, so a delay of seconds sharply weakens the association. The important exception is conditioned taste aversion, which forms even when nausea follows ingestion by hours — a specialized, evolutionarily prepared form of learning that is precisely why nausea-based chemical aversion for alcohol is more robust than its timing would otherwise allow (#ref-howard-2001). The second is extinction. Because the aversion is a learned association, presenting the cue afterward without the aversant erodes it, and on the modern inhibitory-learning view extinction does not delete the original learning but adds a competing one, so an aversion suppressed in treatment can re-emerge in the wider world where the cue is no longer punished (#ref-elkins-1991).

Core Techniques

The procedures differ mainly in the aversant they enlist and whether it is delivered or imagined. Chemical (emetic) aversion, the oldest form, pairs the taste of alcohol with drug-induced nausea and remains the modality with the strongest outcome data (#ref-howard-2001). Electrical (faradic) aversion substitutes a brief, tolerable shock, which is easier to time precisely but, for alcohol, less effective than nausea because shock is not the biologically prepared consequence of drinking. Covert sensitization dispenses with any physical aversant, having the client pair a vividly imagined instance of the behavior with an equally vivid imagined aversive consequence (#ref-cautela-1967). Rapid smoking, developed for smoking cessation, turns the behavior's own properties into the aversant by having the smoker inhale continuously until it becomes noxious (#ref-lichtenstein-1973).

Whatever the aversant, the modern view is that suppression during treatment is not enough. Because the conditioned aversion extinguishes once the cue reappears unpunished, aversion therapy is now understood as at best one component of a broader programme, embedded in the relapse-prevention framework that treats a lapse as a learning opportunity and equips the client with coping skills for the high-risk situations that conditioning alone leaves untouched (#ref-marlatt-1985).

The Evidence Base

The evidence is uneven, and its quality tracks the era in which it was gathered. The early alcohol series reported high abstinence rates — Lemere and Voegtlin's follow-up of more than four thousand patients claimed roughly 60% abstinence at one year — but these were uncontrolled case series from a single institution, without randomization or independent comparison, so their headline figures cannot be read as treatment effects (#ref-lemere-1950). The most rigorous later work used matched comparison groups: a study of inpatient alcoholics treated with chemical aversion found six- and twelve-month abstinence rates significantly above those of matched patients drawn from a treatment registry, the strongest quasi-experimental support the modality has (#ref-smith-1991).

Reviewers who weighed this literature reached cautious and divided conclusions. Ralph Elkins, appraising the chemical-aversion evidence, judged that emetic therapy could produce a genuine conditioned aversion and worthwhile abstinence when properly delivered, while conceding the weakness of much of the supporting design (#ref-elkins-1991). G. Terence Wilson, commenting on the same evidence, was more sceptical, arguing that the methodological limitations — selection effects, the confounding of aversion with the intensive residential programmes around it — left the specific contribution of aversive conditioning unproven (#ref-wilson-1991). Beyond alcohol the record is thinner still: a controlled comparison of three aversion procedures for cocaine dependence found only modest and inconsistent reductions in craving, illustrating how poorly the alcohol results generalize (#ref-bordnick-2004).

Worked Example

Outcome trials for aversion therapy report the proportion abstinent at follow-up, and those proportions translate into a clinically meaningful index exactly as the demonstration above computes. Take representative one-year figures of the order reported for chemical aversion against a comparison group: 60% of the treated group abstinent versus 40% of the comparison group (#ref-smith-1991).

The absolute difference in abstinence is 0.60 − 0.40 = 0.20, or 20 percentage points. The relative rate is 0.60 / 0.40 = 1.5: treated clients are half again as likely to be abstinent. The number needed to treat is the reciprocal of the absolute difference, 1 / 0.20 = 5: on average five clients receive the treatment for one additional person to remain abstinent who would not have under the comparison condition.

None of these figures is exact, and each inherits the design limits of the trials behind it — the comparison groups are matched, not randomized, and the abstinence rates are illustrative round numbers, not pooled estimates. The exercise makes the verdict concrete: where it works, the effect is real and worthwhile but far from curative, and a number needed to treat near five is consistent with a treatment worth offering only as part of a larger programme.

Discussion

Aversive therapy occupies an unusual position in the history of behavior therapy: technically coherent, occasionally effective, and ethically fraught in a way its siblings are not. Its founding insight — that an appetitive behavior can be weakened by conditioning an aversion to its cue — is sound learning theory, and for alcohol use disorder the better-designed studies show a real, if modest, effect. Yet the method's dependence on delivering an unpleasant stimulus makes questions of consent, coercion, and proportionality unavoidable, and its historical application to homosexuality — now recognized as a harmful attempt to change something that was never a disorder — permanently coloured its reputation (#ref-przeworski-2021).

Two technical limits compound the ethical ones. Because the conditioned aversion extinguishes once the cue is encountered without the aversant, its effects are inherently fragile outside the treatment setting, which is why contemporary practice embeds it in relapse-prevention programmes rather than relying on it alone (#ref-marlatt-1985). And because aversion is almost never delivered in isolation from the intensive treatment surrounding it, the specific contribution of the conditioning itself remains hard to isolate, the core of the sceptical reading of the evidence (#ref-wilson-1991). The result is a treatment now used rarely and selectively, retained in a few specialized alcohol programmes rather than as a front-line option.

Current Directions

The most interesting recent work is not clinical expansion but mechanistic clarification. A functional-imaging study of emetic aversion therapy for alcohol use disorder found that successful treatment altered the neural response to alcohol cues, offering the first direct evidence of the brain changes that accompany a conditioned alcohol aversion and grounding a century-old behavioral procedure in identifiable neurobiology (#ref-elkins-2017). This reframes aversion therapy as a specific instance of cue-reactivity modification, connecting it to the wider neuroscience of craving.

Alongside it, the pharmacological landscape that aversion therapy once dominated has shifted. Reviews of novel agents for alcohol use disorder now foreground drugs that reduce craving through neurochemical routes rather than by conditioning revulsion, relegating emetic aversion to a historical and niche role (#ref-burnette-2022). And the reinforcement-based tradition that always stood opposite aversion — strengthening abstinence with rewards rather than punishing use — has accumulated the stronger evidence base: contingency management, in which abstinence earns tangible incentives, is now supported by meta-analytic evidence across substance use disorders, including among patients receiving medication for opioid use disorder (#ref-bolivar-2021). The trajectory of the field points away from aversion and toward incentive- and medication-based approaches, with aversion therapy's likeliest future as a well-understood mechanism rather than a common treatment.

Glossary

Aversant.

The unpleasant stimulus — an emetic drug, an electric shock, or a vividly imagined consequence — paired with the target cue to build an aversion.

Chemical aversion.

Aversion therapy in which an emetic drug pairs drug-induced nausea with the taste of alcohol; the best-supported modality.

Classical conditioning.

Learning in which a neutral stimulus, paired with one that already elicits a response, comes to elicit that response itself; the mechanism aversion therapy exploits.

Conditioned taste aversion.

A prepared form of learning in which a food or drink paired with later nausea becomes aversive, even across a long delay; the basis of emetic aversion.

Contiguity.

The closeness in time between cue and aversant; for most conditioning a short interval is needed, taste aversion being the notable exception.

Contingency management.

The reinforcement-based counterpart to aversion, rewarding verified abstinence with tangible incentives; now the better-evidenced approach.

Covert sensitization.

Cautela's procedure pairing an imagined instance of the unwanted behavior with an imagined aversive consequence, requiring no physical aversant.

Extinction.

The weakening of a conditioned aversion when the cue is met repeatedly without the aversant, the reason the treatment's effects are fragile outside the clinic.

Faradic aversion.

Aversion therapy using a brief electric shock as the aversant; precise in timing but, for alcohol, weaker than nausea.

Number needed to treat.

The average number of clients who must receive a treatment for one additional person to benefit relative to the comparison condition; the reciprocal of the absolute rate difference.

Rapid smoking.

A smoking-cessation aversion procedure in which continuous inhalation makes the act itself noxious, turning the behavior into its own aversant.

Reciprocal inhibition.

Wolpe's principle that one response can suppress an incompatible one; here the conditioned aversion inhibits the appetitive approach.

Relapse prevention.

Marlatt and Gordon's framework that treats a lapse as a learning opportunity and teaches coping skills for high-risk situations, the setting in which aversion is now embedded.

Rescorla-Wagner model.

An associative-learning model in which each pairing changes a cue's strength in proportion to the gap between the outcome's potency and the strength already learned, giving a decelerating acquisition curve.

Key Researchers

Joseph R. Cautela (1927-1999). Psychologist at Boston College who introduced covert sensitization, moving aversive conditioning from delivered aversants to imagined aversive imagery. Britannica

Ralph L. Elkins (1938-2025). Researcher at the Medical College of Georgia and the Augusta VA who ran the longest-sustained experimental programme on chemical aversion for alcohol use disorder, from taste-aversion learning to functional imaging. Obituary

Edward Lichtenstein (living). Emeritus scientist at the Oregon Research Institute whose controlled trials tested rapid smoking as an aversion treatment for smoking cessation. Google Scholar

Stanley Rachman (1934-2021). Professor at the University of British Columbia who co-authored the first book-length analysis of aversion therapy and behaviour disorders before turning to the cognitive account of anxiety. Wikipedia

Walter L. Voegtlin (1904-1975). Physician at the Shadel Sanatorium in Seattle who established the conditioned-reflex treatment of alcoholism, the first large clinical programme of chemical aversion. Wikipedia

G. Terence Wilson (living). Emeritus professor at Rutgers University and a leading critical evaluator of the evidence base and methodology of chemical aversion in alcoholism treatment. Faculty page

Joseph Wolpe (1915-1997). Psychiatrist at Temple University who placed aversion therapy within the reciprocal-inhibition framework of behaviour therapy. Wikipedia

Frequently Asked Questions

What is aversive therapy? Aversive therapy is a behavioral treatment that pairs an unwanted behavior, or the cue that triggers it, with an unpleasant stimulus such as nausea or a mild shock, so the behavior itself comes to evoke discomfort and is suppressed. It is the punishment-based counterpart of reward-based behavior therapies (Rachman & Teasdale, 1969).

What conditions has it been used to treat? Its main and best-supported target is alcohol use disorder, where the taste of alcohol is paired with drug-induced nausea. It has also been applied to smoking, paraphilias, and self-injurious behavior, and historically, and wrongly, to homosexuality (Voegtlin, 1940).

What is covert sensitization? Covert sensitization, developed by Joseph Cautela, is a form of aversion therapy that uses no physical aversant. The client imagines the unwanted behavior and then a vivid, unpleasant consequence, pairing the two in imagination so the behavior acquires an aversive association (Cautela, 1967).

Does aversion therapy actually work? For alcohol use disorder the better-designed studies show a real but modest effect, with matched-comparison trials finding higher abstinence rates in treated patients. Its benefits extinguish once the cue is met without the aversant, so it works best as one part of a larger programme (Smith et al., 1991).

Why do its effects fade over time? Because the aversion is a learned association, encountering the cue outside treatment without the aversant extinguishes it. On the inhibitory-learning view the original learning is not erased but overlaid, so the aversion can re-emerge, which is why relapse prevention is now considered essential (Elkins, 1991).

How is it different from contingency management? The two are mirror images. Aversion therapy weakens a behavior by following it with something unpleasant, while contingency management strengthens abstinence by rewarding it with incentives. Contingency management now has the stronger evidence base across substance use disorders (Bolívar et al., 2021).

Why is aversion therapy ethically controversial? Delivering an aversive stimulus raises unavoidable questions of consent, coercion, and proportionality, and the method was once used to try to change sexual orientation, an application now recognized as ineffective and harmful. That history is the standard cautionary case in discussions of its ethics (Przeworski et al., 2021).

Is aversive therapy still used today? Rarely and selectively. It survives in a few specialized chemical-aversion programmes for alcohol use disorder, while craving-reducing medications and incentive-based treatments have largely displaced it; recent research treats it mainly as a mechanism to understand rather than a common treatment (Burnette et al., 2022).

References