Drug-Induced Psychosis: How Often It Becomes Schizophrenia

Creator: Ryan S. Sultan, MD
Published: 2026-11-12

By Ryan S. Sultan, MD
Psychiatrist · Columbia University Irving Medical Center
Published November 12, 2026 · 5-minute read

About one in four people diagnosed with a substance-induced psychosis are later diagnosed with schizophrenia. A 2020 meta-analysis put the pooled proportion at 25% (95% CI, 18% to 35%), and the type of substance was the primary predictor.¹

In an Ontario cohort, the 3-year risk of a schizophrenia spectrum disorder after an emergency department visit for substance-induced psychosis was 18.5%, compared with 0.1% in the general population.² In the meta-analysis, the rate was highest after cannabis-induced psychosis.¹

Transition rates differ by substance

The meta-analysis included 50 studies of substance-induced, brief, and atypical psychoses, and 25 of them gave substance-specific estimates in 34,244 people.¹

Pooled transition from substance-induced psychosis to schizophrenia
Substance	Transition to schizophrenia
Cannabis	34% (95% CI, 25% to 46%), 6 studies
Hallucinogens	26% (14% to 43%), 3 studies
Amphetamines	22% (14% to 34%), 5 studies
Opioids	12%
Alcohol	10%
Sedatives	9%
All substances	25% (18% to 35%)

Source: Murrie et al., 2020.¹

Transition rates were slightly lower in older cohorts and weren't affected by sex, country, hospital or community setting, diagnostic methods, or duration of follow-up, and the abstract gives no single follow-up period for the pooled figures.¹ The studies report amphetamines as a group, and the abstracts give no separate figure for methamphetamine.

What the national registry studies measured

A Danish study followed all 6,788 people who received a diagnosis of substance-induced psychosis between 1994 and 2014, with follow-up to August 2014. Overall, 32.2% converted to either bipolar or schizophrenia-spectrum disorders, and for cannabis-induced psychosis the figure was 47.4%.³ Both numbers combine schizophrenia and bipolar disorder.

A Finnish study of 18,478 inpatient cases reported schizophrenia-spectrum diagnoses alone. The 8-year cumulative risk was 46% (95% CI, 35% to 57%) after cannabis-induced psychosis, 30% (14% to 46%) after amphetamine-induced psychosis, and 5.0% after alcohol-induced psychosis, which was the most common type.⁴

The Ontario study covered 9,844,497 people aged 14 to 65 with no history of psychosis, including 13,784 with an emergency visit for substance-induced psychosis. Their risk of transition to a schizophrenia spectrum disorder was 163 times that of the general population after adjustment for age and sex. Cannabis carried the highest risk among visits with psychosis, and younger age and male sex were associated with higher risk.²

Emergency visits for substance use without psychosis carried a 3-year risk of 1.4%. Because those visits were far more common, they accounted for more than 3 times the absolute number of transitions (9,969 compared with 3,029).²

In Danish data, half of conversions to schizophrenia came within about three years

In the Danish registry, half of the conversions to schizophrenia occurred within 3.1 years of the substance-induced psychosis, and half of the conversions to bipolar disorder occurred within 4.4 years.³ In the Finnish registry, the majority of conversions occurred during the first 3 years, especially for cannabis-induced psychosis.⁴

The other half came later, and the authors concluded that a long follow-up period is needed to identify the majority of cases.³

How long drug-induced psychosis lasts

A 2016 SAMHSA report reproduces the DSM-IV criteria for substance-induced psychotic disorder and marks DSM-5 criteria B and C as the same. Those criteria don't state a typical duration. Under them, the symptoms develop during, or within a month of, substance intoxication or withdrawal.⁵

The same criteria use "about a month" as a rule-out marker. Symptoms that persist for a substantial period, with about a month given as the example, after acute withdrawal or severe intoxication has ended can count as evidence that a psychotic disorder that isn't substance-induced better explains them.⁵ The 2016 report covers DSM-IV and DSM-5 only.

High-potency cannabis and the odds of a first episode

A case-control study compared 901 patients with first-episode psychosis and 1,237 population controls at 11 sites across Europe and Brazil. Daily cannabis use was associated with higher odds of psychotic disorder compared with never using (adjusted odds ratio 3.2; 95% CI, 2.2 to 4.1). For daily use of high-potency cannabis, defined as THC of 10% or more, the adjusted odds ratio was 4.8 (2.5 to 6.3).⁶

Both figures are odds of first-episode psychosis, a different outcome from conversion of substance-induced psychosis to schizophrenia, which the study didn't measure.

What a later schizophrenia diagnosis means

These studies count people whose diagnosis changed. They can't separate a substance that contributed to schizophrenia from a first episode of schizophrenia that was initially labeled substance-induced. None of the abstracts reports whether stopping the substance after the first episode changed the conversion rate. I've written separately about schizophrenia and psychosis treatment.

Limitations of the evidence

This evidence has limitations. First, every estimate comes from people who received a diagnosis in clinical care, so the figures likely don't apply to brief symptoms that never led to care. Second, the cannabis estimate in the meta-analysis rests on 6 studies, and the registries come from Denmark, Finland, and Ontario. Nevertheless, the estimates agree that the risk is substantial and that much of it falls in the first three years, which supports psychiatric follow-up for several years after a substance-induced psychosis, particularly in young people and after cannabis-induced episodes.

Frequently asked questions

Can drug-induced psychosis turn into schizophrenia?

In a 2020 meta-analysis, the pooled proportion of transition from substance-induced psychosis to schizophrenia was 25% (95% CI, 18% to 35%), with no single follow-up period given. The studies describe a later diagnosis and can't separate a substance that contributed to schizophrenia from a first episode that was initially labeled substance-induced.¹

How likely is cannabis-induced psychosis to become schizophrenia?

Cannabis had the highest pooled rate of transition to schizophrenia in the 2020 meta-analysis, at 34% across 6 studies. In a Finnish inpatient registry, the 8-year cumulative risk of a schizophrenia-spectrum diagnosis after cannabis-induced psychosis was 46%.^{1, 4}

How long does drug-induced psychosis last?

A 2016 SAMHSA report reproduces the DSM-IV criteria and marks DSM-5 criteria B and C as the same, and those criteria don't state a typical duration. Under them, symptoms that persist for a substantial period, with about a month given as the example, after intoxication or withdrawal has ended can count as evidence for a psychotic disorder that isn't substance-induced.⁵

How many years after a drug-induced psychosis can schizophrenia appear?

In a Danish registry, half of the conversions to schizophrenia occurred within 3.1 years of the substance-induced psychosis. The other half came later, and the authors concluded that a long follow-up period is needed.³

Does meth psychosis lead to schizophrenia?

The studies report amphetamines as a group and give no separate figure for methamphetamine. In the 2020 meta-analysis, the pooled transition rate to schizophrenia after amphetamine-induced psychosis was 22%, and in a Finnish inpatient registry the 8-year cumulative risk of a schizophrenia-spectrum diagnosis was 30%.^{1, 4}

In a crisis: call or text 988, or go to the nearest emergency department. A first episode of psychosis belongs in a hospital-based or coordinated specialty care program.

I start and manage clozapine in an outpatient practice in Chelsea, Manhattan. Details are on the schizophrenia care in NYC page. Clozapine should not be stopped, skipped, or restarted without the prescriber. This page is general information and does not replace an evaluation.