Chemsex

Abstract

Chemsex is a form of sexual behavior: the intentional use of psychoactive drugs, typically crystal methamphetamine, GHB/GBL, or mephedrone, to initiate, enhance, and prolong sex among men who have sex with men. Cognitive psychology reframes it not as a moral failing but as a decision made under a chemically altered state, in which drug myopia, the narrowing Steele and Josephs described, pulls attention onto the immediate sexual cue and away from the distal costs of HIV and sexually transmitted infection. The same drugs amplify the visceral arousal that reshapes valuation in the heat of the moment and steepen the rate at which a safer, delayed option is discounted. The influence runs both ways, because repeated drug-sex pairing conditions a learned coupling that reshapes future choice. Three demonstrations trace the narrowing spotlight, stimulant-steepened discounting, and the conditioned link.

Keywords: chemsex, drug myopia, decision-making, delay discounting, sexual risk

Consider two decisions, made by the same person about the same night. The first is made on a quiet afternoon: a plan to use, to set limits, to keep protection within reach. The second is made hours later, several doses into a session, when the drug has pulled the beam of attention tight around the immediate and let everything distal — the risk, the morning, the plan itself — fall into the dark at the edges. Chemsex is where these two decisions diverge most sharply, and the gap between them is not a gap in knowledge or in values but in the state from which each choice was made.

That is why chemsex is a revealing case for cognitive psychology rather than only for public health. Its defining feature is the deliberate recruitment of drugs whose pharmacology acts directly on the machinery of valuation, attention, and self-control: stimulants that intensify and extend arousal, disinhibitors that lower the threshold to act, and all of them narrowing the field of what the decider can hold in view (Giorgetti et al., 2017). The behavior is an unusually pure demonstration of how a chemically altered state reshapes a decision — and, over repeated sessions, of how a decision reshapes the person who will make the next one.

What Chemsex Is

Chemsex, in the sense used by MeSH and by the sexual-health literature, is the intentional use of psychoactive drugs to facilitate, intensify, or prolong sexual activity, a practice documented most extensively among men who have sex with men in urban settings (Bourne et al., 2015; Maxwell et al., 2019). The drugs most characteristically involved — crystal methamphetamine, GHB/GBL, and mephedrone — are chosen precisely for effects that bear on cognition: stimulants heighten arousal, confidence, and stamina; GHB lowers inhibition and distorts the sense of time; and all of them shift what the user attends to and how they weigh it (Giorgetti et al., 2017). What distinguishes chemsex from drug use that merely happens to coincide with sex is the intent: the substances are deployed as instruments of the sexual encounter, which makes the resulting behavior a decision about how to feel and act, taken in advance and then re-taken, differently, in the moment.

The cognitive interest is not in the pharmacology for its own sake but in how predictably these agents bend a decision. Systematic reviews converge on a consistent behavioral picture: chemsex is associated with more sexual partners, longer sessions, and markedly higher rates of condomless sex and of HIV and STI acquisition, even among people who are well informed about those risks and who, interviewed afterwards, can describe exactly the precautions they intended to take (Tomkins et al., 2019; Maxwell et al., 2019). The puzzle this poses is the one cognitive psychology is built to answer. If knowledge is intact and intentions are sincere, the failure must lie not in what the person believes but in how a drug-altered state computes a choice — which cues it attends to, how it values delay, and which control system wins when a strong drive meets a weakened brake.

The drugs characteristic of chemsex are not interchangeable: each bends the decision through a slightly different cognitive route, and matching harm reduction to mechanism begins with naming those routes (Table 1).

Table 1

Common Chemsex Drugs and Their Cognitive Route Into the Sexual Decision

Note. Classes and effects follow the neuropharmacology reviewed by Giorgetti et al. (2017). The routes overlap rather than partition the behavior; most sessions combine agents, compounding their effects on attention, discounting, and inhibition.

How Chemsex Intersects Cognition

Chemsex touches cognitive psychology at three points, the three forward arrows of Figure 1. The drugs narrow attention, so the decision is computed over a shrunken field of cues; they amplify the visceral arousal and steepen the delay discounting that together inflate the immediate option; and across sessions they build an associative link that carries the distortion forward into future choices. Experimental inductions of arousal confirm the behavioral consequence that chemsex drugs intensify: raising arousal increases sexual risk-taking and shifts decision-making toward immediate gratification in both men and women (Skakoon-Sparling et al., 2016). The first two mechanisms describe how a single chemsex decision goes wrong in the moment; the third describes how the moments compound. All three share a logic with the broader self-regulation literature: behavior defaults to the cue-driven, immediate option when the regulatory system that would hold a plan in place is overrun — here, chemically (Metcalfe & Mischel, 1999).

Drug myopia: a narrowing attentional spotlight

The most distinctive cognitive effect of the drugs is not that they manufacture desire but that they subtract the future. Alcohol myopia theory, the foundational account, holds that intoxication narrows attention onto the most immediate and salient cues in a situation, so that behavior comes to be driven by whatever is most present and the inhibiting cues — costs, consequences, conflicting goals — simply fall out of the field of view (Steele & Josephs, 1990). The same mechanism generalises to the stimulants and disinhibitors of chemsex, where it has especially sharp consequences: the immediate sexual reward is maximally salient and central, while the costs that should check it — HIV and STI risk, the possibility of regret, the plan made that afternoon — are distal, abstract, and exactly the kind of cue a narrowed spotlight drops (Giorgetti et al., 2017).

This reframes the risk. It is not that the intoxicated decider weighs the danger and accepts it; it is that the danger is no longer being attended to at all. The choice is made as if the risk cue were absent, because functionally it is. Demo 1 makes this concrete: each cue is weighted by a Gaussian of its psychological distance from the immediate reward, and as the drug load narrows the spotlight the distal risk cues dim toward zero while the proceed-now value wins by subtraction — a reversal driven entirely by what the decider can and cannot see.

Session discounting: a stimulant-steepened rate

If myopia governs which cues enter the decision, delay discounting governs how the surviving cues are weighed across time. Many sexual-risk choices are intertemporal: proceeding immediately without protection is a sooner reward, while pausing to protect defers it. The Sexual Discounting Task measures exactly this trade-off and finds that steeper discounters are far more willing to take the immediate, unprotected option rather than wait, establishing sexual risk as an instance of the general discounting machinery studied in addiction and impulsivity (Johnson & Bruner, 2012). The connection to chemsex is direct, because steep discounting is a signature of addictive behavior across substances: a meta-analysis spanning many drugs of abuse finds that heavier, more problematic use is reliably associated with a higher delay-discount rate (MacKillop et al., 2011).

The stimulant load of a chemsex session pushes the rate up transiently, on top of any trait-level elevation. Reviews of controlled drug-administration studies — the design that can speak to causation rather than mere co-occurrence — find that acute intoxication raises HIV-relevant sexual risk-taking, consistent with a state-driven steepening of the discount rate and a weakening of inhibitory control (Berry & Johnson, 2018). The Worked Example quantifies how small this change needs to be: a rise from a sober rate to a stimulant-steepened one, with the rewards themselves untouched, is enough to flip a sincerely preferred precaution into its opposite. Demo 2 lets the rate and the delay vary around that threshold.

The drug-sex link: how sessions rewire future choice

The first two mechanisms describe a single night. The third describes why chemsex is hard to leave. Each session pairs a powerful drug reward with a powerful sexual one, and that pairing is exactly the condition under which associative learning builds a link between two events. Modelled by the standard Rescorla–Wagner rule, the strength of the drug–sex association grows across paired sessions toward a ceiling. After enough pairings each event becomes a conditioned cue for the other: by incentive-sensitization accounts of addiction, a repeatedly reinforced cue acquires the power to trigger craving and approach, so a sexual context comes to elicit drug wanting and a drug cue to elicit sexual approach (Robinson & Berridge, 1993). This cue reactivity is why people report that sober sex comes to feel incomplete and that the drug reinstates the pull toward sex long after a given session ends, and the steep discounting that heavier use imposes compounds it (MacKillop et al., 2011).

This is the bidirectional claim in its strongest form. A chemsex decision is not only distorted by a drug-altered state; it produces a learned residue that alters the baseline from which the next decision is made — conditioning cue-reactivity, strengthening a habit, and taxing the executive control that would be needed to break the pattern (Metcalfe & Mischel, 1999). Decision-making leads to the behavior, and the behavior rewires decision-making. Demo 3 traces the acquisition curve, showing how a fixed fraction of the remaining gap is closed each session until the coupling approaches its ceiling.

Worked Example

Model a single chemsex-risk choice as a comparison of two discounted values. The safer option — pausing to use protection — delivers a reward of magnitude A = 100 utils but only after a delay of D = 5 (arbitrary time units, the wait to secure and use a condom mid-session). The riskier option — proceeding immediately without protection — delivers a fixed, undelayed reward of R = 60 utils. Using the standard hyperbolic discounting of the delayed reward, its present value is

V_safe = A / (1 + kD)

In a sober state the person discounts shallowly, say k = 0.04. Then

V_safe = 100 / (1 + 0.04 × 5) = 100 / 1.2 = 83.3

Because 83.3 > 60, the discounted value of pausing exceeds the immediate reward, and the person chooses protection. Now let the stimulant load of a session steepen the discount rate to k = 0.20 — well within the transient range a chemsex drug produces. The same delayed reward is now worth

V_safe = 100 / (1 + 0.20 × 5) = 100 / 2 = 50.0

Now 50 < 60, so the ranking reverses and the person proceeds unprotected — without any change in how much they value protection in the abstract, in the delay, or in the reward magnitudes. Only the discount rate moved. The indifference rate, at which the two options are exactly equal, solves 100 / (1 + 5k) = 60, giving k = 0.133: below it protection wins, above it the immediate option does. This is the arithmetic behind the public-health pattern — a modest, drug-driven rise in a single parameter is enough to turn a sincerely intended precaution into condomless sex. Demo 2 recomputes V_safe and the chosen option as k and D are varied.

Discussion

Reading chemsex as a decision problem does two useful things. It replaces a picture of reckless or ill-informed actors with one of ordinary deciders operating under a state that has been chemically engineered to narrow attention, amplify arousal, and steepen discounting — which explains the otherwise puzzling gap between what participants know, what they plan, and what they do (Bourne et al., 2015; Tomkins et al., 2019). And it locates chemsex within the general science of self-regulation rather than treating it as a sealed-off pathology: the same delay discounting, visceral influence, and dual-system control that govern eating, spending, and addiction are the machinery at work here, thrown into unusually sharp relief by the potency of the drugs (MacKillop et al., 2011; Metcalfe & Mischel, 1999). The practical implication follows: because the failure is one of attention and valuation in the moment rather than of knowledge, interventions that restructure the situation — pre-exposure prophylaxis, precommitment, harm-reduction planning done in the cold state — should outperform information alone.

Two cautions keep the account honest. First, much of the human evidence is observational: chemsex is associated with elevated risk, but disentangling the drug's causal contribution from the characteristics of the people who choose chemsex requires the controlled-administration designs that remain few and necessarily limited in the intensity they can ethically reproduce (Berry & Johnson, 2018). Second, the cognitive framing is a description of a valuation contest, not a complete account of why a given person enters chemsex in the first place — motives of intimacy, escape, community, and managing the stigma and trauma that disproportionately burden the populations involved sit outside the discounting arithmetic and are essential to any humane reading of the behavior (Maxwell et al., 2019). Naming the parameter that moved is not the same as naming the life that moved it.

Current Directions

The most active front is the move from association to causation. Because observational data cannot settle whether the drugs cause risky sexual choices or merely accompany them, reviews now prioritise controlled drug-administration studies that manipulate intoxication directly and measure its effect on HIV-relevant decision-making, finding support for a genuine causal contribution rather than pure selection (Berry & Johnson, 2018). A second front is descriptive and epidemiological: systematic reviews are mapping how chemsex practices, drug repertoires, and associated harms vary across settings and populations, which is a precondition for targeting intervention rather than assuming a single profile (Maxwell et al., 2019; Tomkins et al., 2019). A third applies the neuropharmacology of the specific agents to the specific cognitive distortions, asking which drug produces which decision effect — stimulant-driven discounting versus GHB-driven disinhibition and time distortion — so that harm reduction can be matched to mechanism (Giorgetti et al., 2017). The open question uniting them is whether chemsex risk is best reduced by lowering in-the-moment discounting, by strengthening inhibitory control over time, or by decoupling the conditioned drug–sex link that keeps the pattern in place.

Common Misconceptions

People do chemsex because they do not understand the risks.

Knowledge is usually intact. Participants can describe the risks and the precautions they intended, yet choose otherwise once intoxicated — the failure is at the point of choice, where a drug-narrowed spotlight drops the risk cue, not in what the person believes (Bourne et al., 2015; Steele & Josephs, 1990).

The drugs add desire, so the risk is just stronger wanting.

Myopia theory says the drugs mainly subtract the future: they narrow attention onto the immediate cue so that distal costs fall out of view. The decision goes wrong because the risk is no longer attended to, not because desire has simply been turned up (Steele & Josephs, 1990).

Chemsex is a special behavior needing its own psychology.

Its central mechanisms are general ones. Delay discounting, visceral influence, and dual-system self-regulation govern addiction, eating, and spending as much as chemsex; the behavior is a sharp test case for them, not an exception (MacKillop et al., 2011; Metcalfe & Mischel, 1999).

The harm is confined to the session itself.

Repeated drug–sex pairing conditions a learned coupling, so each session alters the baseline for the next: sexual cues come to trigger drug craving and vice versa, which is why the pattern outlasts any single night and resists in-the-moment willpower (MacKillop et al., 2011).

Glossary

Alcohol (drug) myopia.

The theory that intoxication narrows attention onto the most immediate and salient cues, so behavior is driven by what is present and distal inhibiting cues fall out of view (Steele & Josephs, 1990).

Chemsex.

The intentional use of psychoactive drugs to initiate, enhance, or prolong sexual activity, documented chiefly among men who have sex with men; defined here as a decision taken under a drug-altered state.

Cue reactivity.

The conditioned response in which a learned cue — here a drug or a sexual context — triggers craving and approach; the in-the-moment expression of the drug–sex association.

Delay discounting.

The devaluation of a reward as the wait for it grows; the mechanism that lets an immediate unprotected encounter outcompete a safer but delayed one when the discount rate is high.

Dual-process theory.

The account of choice as a contest between a fast, cue-driven, affective system and a slow, deliberative one; chemsex drugs tip the contest toward the former.

Executive function.

The control processes — working memory, inhibition, updating — that hold a goal in place against a strong drive; chemicals that overrun them default behavior to the cue-driven option.

GHB/GBL.

A central-nervous-system depressant pair commonly used in chemsex for disinhibition and euphoria; it distorts the sense of time and carries a narrow margin between an active dose and overdose.

Heat-of-the-moment effect.

The reversal of sexual preference under arousal relative to a calm baseline, together with the failure to anticipate that reversal in advance (Ariely & Loewenstein, 2006).

Hyperbolic discounting.

The form in which discounted value falls as A / (1 + kD) with delay D and rate k; its curvature produces preference reversals when the discount rate changes.

Men who have sex with men (MSM).

The behaviorally defined population among whom chemsex has been most extensively documented; a descriptor of sexual behavior rather than of identity.

Pre-exposure prophylaxis (PrEP).

An antiretroviral taken before exposure to prevent HIV; by removing HIV from the in-the-moment calculation, it acts on the decision rather than on information.

Precommitment.

A choice made in a cold, deliberate state that constrains the later hot-state choice — setting limits or securing protection in advance — a direct countermeasure to drug myopia.

Preference reversal.

The switch in which option is preferred when the discount rate or the arousal state changes, even though the rewards themselves are unchanged; the formal signature of the heat-of-the-moment effect.

Rescorla–Wagner model.

A model of associative learning in which the link between two paired events grows by a fixed fraction of the remaining gap each pairing, V_n+1 = V_n + α(λ − V_n); here it models the forming drug–sex coupling.

Sexual Discounting Task.

A behavioral measure of how a person's likelihood of using a condom falls as the delay to a desired encounter increases, operationalising sexual risk as delay discounting (Johnson & Bruner, 2012).

Visceral factor.

A bodily-motivational state — hunger, craving, sexual arousal — that transiently inflates the value of acting on it and is underweighted when not currently felt; chemsex drugs amplify and prolong it (Loewenstein, 1996).

Key Researchers

Adam Bourne

(La Trobe University, Australian Research Centre in Sex, Health and Society). Led the London qualitative study that defined chemsex in the research literature and set out its HIV and STI risk context among men who have sex with men. Google Scholar · Faculty page

George Loewenstein

(Carnegie Mellon University, Department of Social and Decision Sciences). Originated the visceral-factors account of decision-making, the framework for how drug-amplified arousal narrows valuation onto the immediate during a session. Google Scholar · Wikipedia

James MacKillop

(McMaster University, Peter Boris Centre for Addictions Research). Senior author of the meta-analysis establishing delay discounting as a behavioral signature of addiction across substances, the link that connects chemsex drugs to steepened sexual discounting. ORCID · Faculty page

Walter Mischel

(1930–2018; Columbia University, Department of Psychology). Developed the hot/cool-system analysis of delay of gratification, the framework for how a drug-driven hot system overwhelms cool, deliberative control. Faculty page · Wikipedia

Fabrizio Schifano

(University of Hertfordshire, School of Life and Medical Sciences). Co-authored the neuropharmacology review of chemsex, mapping how the specific agents act on arousal, disinhibition, and risk. Google Scholar · Faculty page

Claude M. Steele

(Stanford University, Department of Psychology). Co-developed alcohol myopia theory, the account of how intoxication narrows attention onto salient cues and lets inhibiting cues fall from view. Faculty page · Wikipedia

Frequently Asked Questions

How does chemsex affect sexual decision-making and risk-taking?

Chemsex drugs act on the decision at three cognitive points. They narrow attention onto the immediate sexual cue (drug myopia), so distal costs such as HIV and STI risk drop out of view; they amplify and prolong the visceral arousal that inflates the value of acting now; and they steepen delay discounting, sinking the value of a safer, delayed option below an immediate one. The result is systematically more risk-taking in the drug-altered state than the same person would choose when sober, even when their knowledge and intentions are unchanged.

Why do people who know the risks still have condomless sex during chemsex?

Because the failure is at the point of choice, not in knowledge. Under drug myopia the risk cue is no longer being attended to, so the decision is computed as if the danger were absent. Participants can describe exactly the precautions they intended and still choose otherwise once intoxicated, which is why information-based prevention alone changes behavior less than expected.

What is drug myopia and how does it change sexual choices?

Drug myopia is the extension of alcohol myopia theory to other intoxicants: the drug narrows the attentional spotlight onto the most immediate, salient cue and lets distal, abstract cues fade. In a sexual situation the immediate reward is maximally salient while the costs are distal, so a narrowed spotlight drops the costs and the immediate option wins by subtraction rather than by any increase in desire.

How does chemsex relate to delay discounting and impulsivity?

Sexual risk can be modelled as delay discounting: an immediate unprotected encounter is a sooner reward and pausing to protect defers it. Steeper discounters more readily take the immediate option, and heavier drug use is reliably linked to steeper discounting, so a stimulant load pushes the rate up transiently. A modest rise in the discount rate, with the rewards unchanged, is enough to flip a preferred precaution into condomless sex.

Does chemsex change decision-making even outside the session?

Yes. Repeated pairing of the drug with sex conditions an associative link, so that over time a sexual cue triggers drug craving and a drug cue triggers sexual approach. This learned coupling alters the baseline from which future decisions are made, which is why sober sex can come to feel incomplete and why the pattern resists in-the-moment willpower.

Is chemsex risk about the drug or about the person?

Both, and separating them is an active research question. Observational data show chemsex is associated with elevated risk, but controlled drug-administration studies — which manipulate intoxication directly — indicate a genuine causal contribution from the drugs on top of any trait-level differences among the people who choose chemsex. Trait-level steep discounting and state-level intoxication compound rather than exclude each other.

How does sexual arousal interact with drugs in the heat of the moment?

Sexual arousal is a visceral factor that inflates the momentary value of acting and is underweighted when absent, producing a gap between the cold plan and the hot choice. Chemsex drugs amplify and prolong that arousal while simultaneously narrowing attention and steepening discounting, so the three effects reinforce one another and widen the cold–hot gap beyond what arousal alone would produce.

What do cognitive models suggest about reducing chemsex harm?

Because the breakdown is one of attention and valuation in the moment rather than of knowledge, strategies that act on the situation tend to outperform information alone. Pre-exposure prophylaxis removes HIV risk from the discounting calculation entirely; precommitment and harm-reduction plans made in the cold state bind the later hot-state choice; and decoupling the conditioned drug–sex link addresses the learned residue that keeps the pattern in place.

References

Ariely, D., & Loewenstein, G. (2006). The heat of the moment: The effect of sexual arousal on sexual decision making. Journal of Behavioral Decision Making, 19(2), 87–98. https://doi.org/10.1002/bdm.501

Berry, M. S., & Johnson, M. W. (2018). Does being drunk or high cause HIV sexual risk behavior? A systematic review and meta-analysis of drug administration studies. Pharmacology Biochemistry and Behavior, 164, 125–138. https://doi.org/10.1016/j.pbb.2017.08.009

Bourne, A., Reid, D., Hickson, F., Torres-Rueda, S., & Weatherburn, P. (2015). Illicit drug use in sexual settings (‘chemsex’) and HIV/STI transmission risk behaviour among gay men in south London: Findings from a qualitative study. Sexually Transmitted Infections, 91(8), 564–568. https://doi.org/10.1136/sextrans-2015-052052

Giorgetti, R., Tagliabracci, A., Schifano, F., Zaami, S., Marinelli, E., & Busardò, F. P. (2017). When “chems” meet sex: A rising phenomenon called “chemsex”. Current Neuropharmacology, 15(5), 762–770. https://doi.org/10.2174/1570159X15666161117151148

Johnson, M. W., & Bruner, N. R. (2012). The Sexual Discounting Task: HIV risk behavior and the discounting of delayed sexual rewards in cocaine dependence. Drug and Alcohol Dependence, 123(1–3), 15–21. https://doi.org/10.1016/j.drugalcdep.2011.09.032

Loewenstein, G. (1996). Out of control: Visceral influences on behavior. Organizational Behavior and Human Decision Processes, 65(3), 272–292. https://doi.org/10.1006/obhd.1996.0028

MacKillop, J., Amlung, M. T., Few, L. R., Ray, L. A., Sweet, L. H., & Munafò, M. R. (2011). Delayed reward discounting and addictive behavior: A meta-analysis. Psychopharmacology, 216(3), 305–321. https://doi.org/10.1007/s00213-011-2229-0

Maxwell, S., Shahmanesh, M., & Gafos, M. (2019). Chemsex behaviours among men who have sex with men: A systematic review of the literature. International Journal of Drug Policy, 63, 74–89. https://doi.org/10.1016/j.drugpo.2018.11.014

Metcalfe, J., & Mischel, W. (1999). A hot/cool-system analysis of delay of gratification: Dynamics of willpower. Psychological Review, 106(1), 3–19. https://doi.org/10.1037/0033-295X.106.1.3

Robinson, T. E., & Berridge, K. C. (1993). The neural basis of drug craving: An incentive-sensitization theory of addiction. Brain Research Reviews, 18(3), 247–291. https://doi.org/10.1016/0165-0173(93)90013-P

Skakoon-Sparling, S., Cramer, K. M., & Shuper, P. A. (2016). The impact of sexual arousal on sexual risk-taking and decision-making in men and women. Archives of Sexual Behavior, 45(1), 33–42. https://doi.org/10.1007/s10508-015-0589-y

Steele, C. M., & Josephs, R. A. (1990). Alcohol myopia: Its prized and dangerous effects. American Psychologist, 45(8), 921–933. https://doi.org/10.1037/0003-066X.45.8.921

Tomkins, A., George, R., & Kliner, M. (2019). Sexualised drug taking among men who have sex with men: A systematic review. Perspectives in Public Health, 139(1), 23–33. https://doi.org/10.1177/1757913918778872