Is Ketamine Addictive? What the Evidence Shows

Is Ketamine Therapy Addictive?

Ketamine can be addictive, but addiction appears to be rare when ketamine is prescribed and monitored as part of a clinical protocol. Across published cohorts totaling several thousand patients treated for depression and related conditions, ketamine addiction has been documented in a small number of individual cases, most often in people with a prior history of substance misuse.

How Often It Actually Happens

Most of the useful data comes from patients treated for depression, which is also where this question tends to come up.

Smith-Apeldoorn and colleagues (2022) reviewed 1,495 patients receiving maintenance ketamine treatment and identified one case of ketamine addiction, in a patient who already had a history of substance misuse. Across the open-label trials they examined, they found no reports of craving or drug-seeking behavior.¹

A separate group of 6,630 patients receiving repeated ketamine for depression offers the same picture from a different angle. Forty-seven of them stopped treatment because of adverse effects, and the adverse effects that prompted those discontinuations "did not include addiction or cognitive deficits."⁹

Reviewing 30 studies specifically for craving, drug liking, tolerance, withdrawal, dose escalation, and illicit use, Vasudeva and colleagues (2026) found that clinical studies "consistently reported minimal evidence of craving, dose escalation, misuse, or illicit use when ketamine or esketamine was administered in controlled, clinically supervised settings."¹⁰

At-home sublingual treatment has been measured as well. In a study of more than 11,000 patients, Mathai and colleagues (2024) recorded cravings as a reported side effect in 0.17% to 0.36% of patients responding at each check-in. The authors flagged one thing worth repeating: among the adverse events they observed, drug cravings was the one that did not match the established risks of commonly prescribed antidepressants.¹¹

None of which makes the risk zero. Roelandt and colleagues (2026) published the case of a 25-year-old woman with no history of addiction who received a single supervised dose of intranasal ketamine in a trial, then sought out illicit ketamine and escalated to daily use within weeks. Her clinicians described it as evidence of ketamine's addictive risk "even in controlled settings."²

Is Ketamine Physically Addictive?

Ketamine's dependence risk is believed to be primarily psychological rather than physical. When it does develop, people are more likely to describe wanting the experience and the relief than avoiding a physical withdrawal.

In the largest survey to date of people living with ketamine use disorder, the abstinence syndrome participants described was emotional rather than physical: cravings in 71%, low mood in 62%, anxiety in 59%, and irritability in 45%.⁸ The DEA's own classification reflects the same asymmetry, describing an abuse potential that "may lead to moderate or low physical dependence or high psychological dependence."³

How Ketamine's Addiction Potential Compares

Ketamine has been a Schedule III controlled substance since 1999, a category the DEA defines as carrying moderate to low potential for dependence.³

The company it keeps in that tier is instructive: Schedule III also contains buprenorphine, prescribed as Suboxone to treat opioid use disorder. Ketamine sits a tier below fentanyl, cocaine, oxycodone, and Adderall, and a tier above tramadol and zolpidem.³

Independent ranking lands in a similar place. When 101 German addiction medicine physicians scored 33 psychoactive substances across five dimensions of harm, ketamine placed 13th of 29, below crack, methamphetamine, heroin, alcohol, cocaine, and amphetamines (Bonnet et al., 2020). That ranking reflects expert judgment on overall harm rather than measured dependence, but it indicates where clinicians who treat addiction place ketamine relative to substances they see daily.¹²

A consortium of mood disorder physicians who prescribe ketamine, including sublingual and intranasal formulations for home use, examined the question directly and found "no clear evidence that addiction potential of ketamine is more serious than other drugs we prescribe with due caution in psychiatry such as stimulants or sedatives" (Swainson et al., 2022).⁴

Therapeutic Use vs. Recreational Misuse: Why the Risk Is Different

Almost everything the public knows about ketamine addiction comes from recreational use. That matters, because recreational and therapeutic ketamine differ on nearly every variable that influences whether a drug becomes habit-forming: how much, how often, where, and with whom watching.

_{Sources for this table: Fang et al. 2026,¹⁵ Harding et al. 2025,⁸ Vasudeva et al. 2026,¹⁰ and McIntyre et al. 2021 as reported in Vasudeva.¹⁰}

How Supervised Dosing Changes the Risk

The research converges on the same handful of variables, and it does so from different directions.

Reviewing where misuse actually appeared across 30 studies, Vasudeva and colleagues (2026) found it "was most often associated with contexts involving unsupervised dosing or dose escalation outside structured treatment protocols."¹⁰ Examining the small number of published cases of addiction in patients treated for depression, Swainson and colleagues (2022) reached a similar conclusion: "poor monitoring of ketamine prescribing and inappropriate patient selection were thought to contribute to the misuse."⁴

Investigators running the addiction trials named the safeguards explicitly. Dakwar and colleagues (2020) credited "careful patient selection, preparation, monitoring, and postprocedure support."⁶ The team behind a trial of ketamine for alcohol use disorder described the risks as minimized when treatment is given to patients who are "appropriately screened, supervised, and followed up."¹³

What Ketamine Does in the Brain

There also appears to be a mechanistic reason the therapeutic signal stays low, and it may distinguish ketamine from drugs with higher addiction liability.

Working in mice, Simmler and colleagues (2022) compared ketamine directly against cocaine. Both raised dopamine in the brain's reward center, but cocaine's rise persisted while ketamine's was a brief burst that shut itself off: ketamine briefly releases the brake on dopamine release, and the resulting surge is then quenched by the brain's own feedback receptors. When researchers blocked those receptors, ketamine produced prolonged dopamine release instead. Deleting the receptor subunit ketamine binds to abolished the ketamine burst while barely affecting cocaine or fentanyl, making this a ketamine-specific mechanism.¹⁴

The behavioral results tracked the biology. Addiction is defined by compulsive use despite negative consequences, and ketamine failed that test: when extra effort was required, mice worked for cocaine but not for ketamine, and they would not take ketamine when doing so carried a negative consequence. The same animals did voluntarily consume ketamine and cocaine in similar amounts, so ketamine is genuinely rewarding. What it did not produce was the compulsive pattern.¹⁴

This is animal research, and it explains rather than establishes the clinical picture described above.

Using Ketamine to Treat Other Addictions: What the Research Shows

Early evidence suggests ketamine may improve abstinence, at least in the short term. A 2026 systematic review pooling 15 randomized trials and 798 participants across alcohol, cocaine, opioid, and tobacco use disorders found that ketamine improved abstinence in the first month after treatment, with moderate-certainty evidence. Researchers attribute the effect to ketamine's action on the learned links between cue and reward that sustain addiction. Beyond one month the effect ran in the same direction but was no longer statistically significant, so the honest word here is promising rather than proven (Fang et al., 2026).¹⁵

Why Researchers Think It Might Help

Repeated substance use strengthens the links between cues and reward, and those connections are believed to be part of what drives conditioned craving and relapse. Ketamine's rapid effect on synaptic plasticity may work against that consolidation, and this rationale is independent of any antidepressant effect (Hakami, 2026).¹⁶ A related pathway is more specific: memories become briefly unstable when they are recalled, and in one trial a single ketamine infusion given after participants deliberately recalled alcohol-related memories reduced drinking and craving for up to nine months, compared with recall alone or ketamine without recall.¹⁶

Alcohol Use Disorder

Dakwar and colleagues (2020) gave 40 adults with alcohol dependence a single ketamine infusion or midazolam as an active control during five weeks of motivational enhancement therapy. The ketamine group had more abstinent days, took longer to relapse, and had fewer heavy drinking days.⁶ Grabski and colleagues (2022) randomized 96 patients with severe alcohol use disorder and found more days abstinent at six months, with the largest effect among those who also received relapse-prevention therapy.⁷ Results are not uniform: a trial in patients with both alcohol use disorder and major depression found no significant difference in abstinence, craving, or consumption.¹⁵

Cocaine and Opioid Use Disorders

The cocaine evidence is the best-controlled in this literature. Dakwar and colleagues (2019) randomized 55 cocaine-dependent adults to ketamine or midazolam alongside mindfulness-based relapse prevention, and reported higher abstinence, a 53% lower risk of relapse, and substantially lower craving.⁵ Opioid data is thinner: one early trial found ketamine relieved acute withdrawal but had limited effect on longer-term abstinence.¹⁵

The Paradox: Can Treatment Create a New Dependence?

The trials above were built to look for exactly this.

Each enrolled people with an active addiction while screening out anyone with a history of ketamine misuse specifically, which makes them unusually informative. They test whether ketamine dependence develops from scratch in people who have already shown a vulnerability to addiction.

Both Dakwar trials tracked it directly, monitoring for "emergence of new drug misuse (ketamine, opioids, or benzodiazepines)" through self-report, urine toxicology, and psychiatric assessment. Neither found any instances, in treatment or control groups.⁵⁻⁶

The most informative single finding came from Grabski and colleagues. Six participants used ketamine on one occasion during follow-up. Three had received ketamine during the trial and three had received placebo, and all six had used ketamine recreationally before enrolling.⁷ Prior history tracked with who used it. Receiving ketamine as treatment did not.

Dakwar and colleagues concluded the risk "can be effectively managed even when this agent is given to substance users."⁶ One caveat belongs with that: participants were told they might receive any of several compounds, so they did not know they were getting ketamine. Real patients do, and the investigators raised this themselves as a reason their findings may understate ordinary practice.⁵

Who Is Most at Risk, and How Supervised Programs Reduce It

Risk is not evenly distributed. The clearest signal in the research is prior ketamine use, as the alcohol trial above showed, and the specialists who prescribe these formulations have turned that into a screening profile.

Swainson and colleagues (2022) list what they look for before prescribing ketamine for less supervised use: no history of drug misuse, no history of drug diversion, a medically suitable profile including controlled blood pressure, reliability in attending follow-up, and a willingness to report side effects.⁴

Where the disorder actually begins is instructive as well. In the largest survey of people living with ketamine use disorder, 72% first obtained ketamine from friends and 14% from dealers. Four of the 274 respondents first obtained it through a clinical prescription (Harding et al., 2025).⁸

Screening, Monitoring, and Treatment Structure

Swainson's group also published practical guidance for prescribing ketamine for home use, and it centers on controlling supply and watching for specific behaviors: prescribe limited quantities with limited refills, then monitor for lost prescriptions, requests for early refills, and requests for dose escalation despite a stable psychiatric picture.⁴

Published protocols apply versions of this. In Mindbloom's subcutaneous program, clinicians reviewed state Prescription Drug Monitoring Program records before every order and limited prescriptions to the minimum quantity required for the treatment phase, a peer treatment monitor was required to be present during sessions, and longitudinal surveys screened for emerging dependency (Parks et al., 2026).¹⁷

Signs of Ketamine Use Disorder

The behaviors clinicians watch for are the same ones worth noticing in yourself: needing more to get the same effect, using between scheduled sessions, seeking ketamine outside the prescription, or continuing to use despite it causing problems. Any of these is worth raising with a prescriber.

What This Means If You're Considering Ketamine Therapy

Ketamine's addictive potential is real, and it is not the thing most likely to go wrong in supervised treatment. The cases that appear in the literature cluster around unsupervised dosing, escalation outside a protocol, and prior ketamine use, rather than around treatment itself.

The honest limit is that no one has run the study that would settle the question completely. No trial has followed people with an addiction over years, in a take-home model, while systematically measuring ketamine craving, liking, and dose-seeking. Until that exists, the reasonable position is the one the specialists take: screen properly, control the supply, watch for specific behaviors, and treat the risk as manageable rather than either trivial or disqualifying.

If you are weighing ketamine therapy and have a history of substance use, that history is worth raising directly with a prescribing clinician. It changes how a program should be structured, and sometimes whether it is appropriate at all.