What Drugs Cause Psychosis and How Long It Lasts

A wide range of substances can trigger psychosis, from street drugs like methamphetamine and cannabis to prescription medications like corticosteroids and ADHD stimulants. Duration is equally varied: some drug-induced psychotic episodes clear within hours of the substance leaving the body, while others persist for weeks or, in a meaningful minority of cases, mark the beginning of a longer-term psychiatric disorder. The clinical rule of thumb is that substance-induced psychosis should resolve within about 30 days of sobriety, but that timeline is a rough guideline, not a guarantee.

Cannabis and High-Potency THC

Cannabis is the substance most commonly linked to psychosis in recent research, and potency matters enormously. A study comparing people experiencing a first psychotic episode with healthy controls found that those in the psychosis group were far more likely to have used high-potency cannabis: about 78% of the psychosis group used potent strains compared with 37% of the control group. The association was strong enough to suggest that THC itself is the ingredient driving the elevated risk, with higher potency, longer duration of use, and more frequent use all independently raising the odds.

What makes cannabis-induced psychosis particularly concerning is what happens afterward. A large register-based study found that roughly 47% of people hospitalized for cannabis-induced psychosis eventually received a diagnosis of schizophrenia or bipolar disorder, the highest conversion rate among all substance-induced psychoses studied.¹ A more recent meta-analysis put the figure somewhat lower, with about 20% later receiving a schizophrenia spectrum diagnosis and 5% a bipolar diagnosis.² The discrepancy likely reflects differences in how conversion was defined and how long people were followed. Either way, cannabis-induced psychosis is not always a one-off event. A nationwide Korean study tracking patients over roughly a decade found that cumulative conversion rates to a psychotic disorder climbed steadily over time, reaching about 29% at ten years for all first-episode psychoses combined.³

For those whose psychosis does not convert, symptoms typically resolve within days to a few weeks after stopping cannabis use. But the people for whom it does convert often cannot be identified in advance by symptoms alone, which is part of what makes the condition so unpredictable.

Methamphetamine and Amphetamines

Methamphetamine is one of the most reliable psychosis triggers among recreational drugs. The mechanism involves a massive flood of dopamine, but the damage goes beyond a temporary chemical imbalance. Longer use appears to physically reduce the density of dopamine transporters in the brain, and that reduction can persist even after a person stops using the drug.⁴ The neurodegeneration seems to involve a cascade of oxidative stress and inflammation in the brain’s dopamine pathways.⁵

What sets methamphetamine apart from many other substances is the phenomenon of sensitization. With repeated use, the threshold for triggering psychosis drops. Researchers have described a progression from normal mental function to a “prepsychotic” state to full-blown psychosis, along with an enhanced vulnerability to relapse that can last for years.⁶ Even after long-term abstinence, mild psychosocial stressors can trigger “flashback” psychotic episodes. In one study, people who had previously experienced methamphetamine psychosis and then encountered stressful events showed spikes in stress hormones and dopamine metabolites that re-triggered psychotic symptoms without any new drug exposure.⁷

In practical terms, the duration of methamphetamine psychosis ranges widely. An acute episode during or just after a binge may resolve in a few days. But for heavy, long-term users, psychotic symptoms can persist for weeks or months after the last dose. And the sensitization effect means the vulnerability may never fully disappear.

Cocaine

Cocaine-induced psychosis tends to look different from schizophrenia-like presentations. Its hallmark is paranoia, which shows up in a strikingly high proportion of users. Research has found paranoia occurring in roughly 68% to 84% of people using cocaine, and violent behavior occurs in as many as 55% of those with cocaine-induced psychiatric symptoms.⁸ The paranoia can be intense enough that people barricade themselves indoors or become aggressive toward others they perceive as threats.

The good news, relatively speaking, is that cocaine psychosis tends to be shorter-lived than methamphetamine psychosis. Because cocaine clears the body faster, symptoms usually begin resolving within hours to a couple of days once use stops. Chronic heavy users, however, may develop a more persistent paranoid state that takes longer to fully clear.

Prescription Stimulants for ADHD

Prescription amphetamines and methylphenidate, the two main classes of ADHD medication, can also cause psychosis, though the absolute risk is low. A large study of people with ADHD starting stimulant treatment found psychosis episodes occurred at a rate of about 2.4 per 1,000 person-years. Amphetamine-based medications carried a higher risk than methylphenidate, with roughly 60% greater odds of a psychotic episode.⁹ A more recent meta-analysis confirmed the pattern, finding that amphetamine exposure was associated with about 57% higher odds of psychosis compared with methylphenidate.¹⁰

These episodes tend to resolve when the medication is stopped or the dose is reduced. Still, the distinction between “the drug caused it” and “the drug unmasked an underlying vulnerability” is difficult to make in any individual case, which is one reason follow-up matters even when symptoms clear quickly.

Corticosteroids

Steroids like prednisone and dexamethasone, prescribed for inflammation and autoimmune conditions, are an underappreciated cause of psychosis. The risk scales with dose: among patients taking the equivalent of 40 mg per day of prednisone or less, roughly 1.3% developed psychiatric disturbances, but that figure jumped to about 18% in those taking 80 mg per day or more.¹¹ Most psychiatric side effects show up within the first two weeks, with a median onset around 11 to 12 days, though onset can happen as quickly as the first day of treatment.¹²

The reassuring aspect of steroid-induced psychosis is that it reliably resolves once the steroid is tapered or discontinued. In one review, 90% of cases resolved within six weeks. Patients treated with both dose reduction and an antipsychotic medication returned to their psychological baseline, though the exact timeline varied significantly from case to case.¹³ Interestingly, the dose of steroid predicts whether psychosis will develop, but it does not predict how severe the psychosis will be or how long it will take to clear.

Dissociatives and PCP

PCP (phencyclidine) and ketamine are both NMDA receptor antagonists, and both can produce psychotic symptoms that look remarkably like schizophrenia, including flattened emotions, disorganized thinking, depersonalization, and catatonic states.¹⁴ In fact, PCP-induced psychosis was historically one of the main laboratory models for studying schizophrenia, because the resemblance extends to the “negative” symptoms (social withdrawal, blunted affect) that other drug models tend to miss.

In chronic PCP or ketamine users, psychotic symptoms can persist for prolonged periods after the last dose, and long-term addicts often develop lasting memory problems.¹⁵ For occasional users, acute ketamine-induced psychotic symptoms are generally brief, clearing as the drug is metabolized over a matter of hours. PCP tends to produce longer episodes because it stays in the body longer and accumulates in fat tissue.

Synthetic Cannabinoids and Cathinones

Synthetic cannabinoids (sold under names like “Spice” or “K2”) and synthetic cathinones (“bath salts”) deserve their own category because their psychosis profiles are notably more severe than their natural-plant counterparts. In one comparative study, hospitalizations for psychosis caused by synthetic cannabinoids lasted nearly twice as long as those for natural cannabis, averaging about 43 days versus 23 days.¹⁶ A case series found that about 30% of patients experienced prolonged psychotic episodes lasting more than five months.¹⁷

Part of what makes these substances so unpredictable is that their chemical composition varies from batch to batch. Users often have no idea what compound or dose they are taking, which makes it impossible to anticipate psychotic reactions or advise on expected duration. Clinicians treating synthetic cannabinoid psychosis sometimes find themselves managing episodes that look more like an unknown toxicological exposure than a standard drug reaction.

Classic Psychedelics and HPPD

LSD, psilocybin, and similar serotonergic psychedelics occupy a complicated position. Frank psychosis during the acute experience (a “bad trip” with loss of reality testing) is relatively uncommon in supervised settings, but a distinct condition called hallucinogen persisting perception disorder (HPPD) can follow use. HPPD involves lingering visual disturbances like intensified colors, afterimages, or movement in the periphery of vision.¹⁸

The tricky part is separating subclinical visual oddities from a genuine clinical disorder. A prospective study found that about a third of participants reported at least one HPPD-like visual symptom four weeks after psychedelic use, but fewer than 1% found the effects distressing. Clinically impairing HPPD appears to be uncommon, and its risk after supervised therapeutic use remains uncertain.¹⁹ People with a psychiatric history, younger age, and high trait absorption seem to be at greater risk for persistent symptoms.

Alcohol Withdrawal

Alcohol itself rarely causes psychosis during intoxication, but stopping heavy drinking abruptly can trigger it. Delirium tremens typically begins 48 to 72 hours after the last drink and represents a severe withdrawal state that can include vivid hallucinations, paranoid delusions, and profound confusion. This type of withdrawal psychosis can last up to two weeks if untreated.²⁰ It is also a medical emergency: delirium tremens carries a meaningful risk of death without proper hospital management, which distinguishes it from most other drug-induced psychoses.

Anticholinergic Drugs and Dopamine Agonists

Two other prescription categories round out the less-discussed psychosis triggers. Anticholinergic agents, which include certain antihistamines, bladder medications, and older antidepressants, can produce psychosis as part of a broader toxicity syndrome that includes confusion, visual hallucinations, and agitation.²¹ These episodes are typically dose-dependent and resolve once the offending drug is removed.

Dopamine agonists, drugs prescribed for Parkinson’s disease and certain hormonal conditions, can also induce psychosis by overstimulating dopamine receptors. A case report documented psychotic symptoms emerging in a patient with no psychiatric history after a dose increase of cabergoline, a dopamine agonist used for a pituitary condition.²² This risk is better recognized in Parkinson’s patients but may be underappreciated in other populations.

Who Is Most Vulnerable

Not everyone who uses the same substance at the same dose will develop psychosis, and genetics is a major reason why. Research using polygenic risk scores has found that people with a higher inherited predisposition to schizophrenia are more likely to use cannabis, use it heavily, and develop cannabis use disorder. The genetic overlap goes both directions: the same genetic variants that increase schizophrenia risk also seem to increase the drive toward cannabis use.²³ This makes it difficult to untangle whether cannabis “causes” psychosis or whether people genetically prone to psychosis are also drawn to cannabis, or both.

Beyond genetics, environmental stressors magnify the risk. Sleep deprivation alone can produce psychotic-like symptoms. Research has consistently shown that extended wakefulness reliably produces perceptual distortions and hallucinations in a majority of participants, with one review finding positive results in 95% of the studies examined.²⁴ Stimulant binges often involve days without sleep, so the psychosis that follows may reflect the combined effects of the drug and severe sleep loss rather than the drug alone.

How Treatment Works and How Long Recovery Takes

The clinical expectation is that substance-induced psychosis resolves within about 30 days of sobriety, but people with this condition are at higher risk for developing severe addiction afterward.²⁵ Antipsychotic medications are typically the first-line treatment, started at low doses and gradually increased. The goal is stabilization followed by slow tapering once the patient is doing well.

For cannabis-induced psychosis specifically, a large real-world study found that antipsychotic use reduced the risk of being rehospitalized for psychosis by about 25%. Long-acting injectable formulations of aripiprazole and olanzapine were associated with the strongest protective effects, reducing rehospitalization risk by roughly 70%.²⁶ Despite the “substance-induced” label suggesting a temporary problem, a substantial proportion of patients end up on antipsychotics for a year or more. One nationwide study found that about 19% of people with substance-induced psychosis were still using antipsychotics three years after their first diagnosis, compared with 45% of those diagnosed with a primary psychotic disorder.²⁷

When the Brain Looks Different

Neuroimaging research has found that people with substance-induced psychosis show brain structural changes that overlap substantially with those seen in schizophrenia, including reductions in gray matter volume. But there are substance-specific patterns as well: methamphetamine users, for instance, show more pronounced shrinkage in the amygdala than people with primary psychosis. Overall, the structural deficits tend to be more severe and widespread in schizophrenia than in purely substance-induced cases, which may explain why many substance-induced cases eventually clear while primary psychotic disorders generally do not.²⁸

Legal Implications of Drug-Induced Psychosis

In the legal system, drug-induced psychosis sits in an awkward gray area. Psychosis from acute intoxication has never been accepted as the basis for an insanity defense in U.S. courts. If you voluntarily take a substance and become psychotic, the law generally holds you responsible for your actions. However, most jurisdictions have carved out an exception for what is sometimes called “settled insanity,” where the long-term downstream effects of substance use have caused lasting mental impairment. In those cases, the psychosis is treated more like a chronic mental illness than a temporary drug reaction, and it can serve as grounds for an insanity plea.²⁹ The distinction between “temporary intoxication” and “permanent brain damage” is the hinge on which these cases turn, and it is often difficult for clinicians or courts to determine which category a given person falls into.