Skip to content
DE

What California sober means, and the health trade-offs involved

Clinical review pending

What California sober means, what research shows about replacing alcohol with cannabis, and the health trade-offs to consider.

A clear lens and blue woven canopy hover above a yellow and ivory capsule.

Could replacing alcohol with cannabis reduce alcohol-related harm? “California sober” usually means giving up alcohol while continuing to use cannabis. Whether that produces a better overall outcome depends on how much and how often someone uses each substance, their health history, and whether a substance use disorder is present. It is not an established treatment for alcohol use disorder.

What does “California sober” mean?

People who describe themselves as California sober generally avoid alcohol and most other drugs but continue using cannabis. Some also include occasional psychedelic use. There is no agreed definition, so two people using the label may have very different patterns of consumption.

The term gained widespread attention in 2021 through Demi Lovato’s public discussion of it. The underlying idea is older: cannabis substitution was studied in 2009 and explicitly discussed as harm reduction in a peer-reviewed paper in 2015 (Reiman, 2009; Lau et al., 2015).

California sober is distinct from complete abstinence and from the broader process of recovery from a substance use disorder. “Sober curious,” meaning actively reconsidering one’s relationship with alcohol, and harm reduction also have their own cultural and clinical meanings. A label alone does not tell us whether someone is recovering, receiving treatment, or experiencing fewer harms.

People drawn to the approach commonly hope to:

  • Drink much less alcohol or stop drinking altogether.
  • Reduce the health risks they associate with continued drinking.
  • Continue using a substance socially or recreationally while avoiding alcohol-related consequences such as hangovers, liver injury, or accidents.
  • Find an approach that feels more flexible than an abstinence-focused model of recovery.

These are motivations, not established benefits. Cannabis can also impair judgment and coordination and contribute to accidents. Replacing one substance changes the risks rather than automatically removing them.

What research on cannabis as an alcohol substitute shows

Three kinds of evidence matter: studies of substitution, trials of addiction treatment, and research on cannabis-related risks. They answer different questions. Does drinking decrease? What additional risks appear? Does the approach help people recover from alcohol use disorder? Those questions cannot be collapsed into one verdict about a loosely defined lifestyle label.

Does substitution substantially reduce drinking? Mixed evidence, including an acute trial

A systematic review of human and animal studies found mixed patterns. Cannabis can substitute for alcohol, but it can also be used alongside it. The population, setting, and reasons for use influence which pattern occurs.

Observational studies of medical cannabis patients or authorized users suggest that some people drink less after using cannabis. These groups selected cannabis themselves, however, so their experiences do not show what would happen if other people were advised to switch. Other research documents complementary use, in which consumption of both substances rises (substitution study; authorized-patient survey; systematic review; review of co-use).

A newer placebo-controlled crossover trial adds experimental evidence. Published online in 2025 and in a 2026 journal issue, it randomized 157 people who drank heavily and already used cannabis at least twice weekly; 138 completed at least two sessions. After smoking cannabis containing 3.1% or 7.2% THC, participants consumed 19% or 27% less alcohol, respectively, during a laboratory task than after placebo. The main craving questionnaire did not show a significant effect. This was an acute experiment after overnight cannabis abstinence, not a trial of sustained substitution, recovery, or reduced long-term harm. It should not be extrapolated to people who do not already use cannabis (Metrik et al., 2026).

Is cannabis less harmful than alcohol? The comparison depends on exposure and health

A useful comparison considers dose, duration, frequency, product, and the person’s health. Heavy alcohol use has well-described effects on the liver and cardiovascular system and increases cancer risk. Heavy cannabis use has a different set of risks. Neither “natural” nor “less harmful than something else” means harmless.

Research has found cognitive differences in adolescents and young adults who use cannabis frequently. In that meta-analysis, the average association was small and was not statistically significant in studies requiring more than 72 hours of abstinence; it should not be described as proof of permanent damage in every young user. A clinical review discusses heavy use, dependence, and brain function. Other work examines cognitive outcomes of long-term regular adult use and associations with mood disorders. Long-term use has also been associated with lower cognitive reserves in midlife. These findings differ in design and cannot all establish causation.

Cannabis use disorder is a recognized condition. A comparison of two substances therefore needs to consider the possibility that alcohol-related problems decrease while cannabis-related problems emerge or worsen.

Is cannabis an evidence-based treatment for alcohol use disorder?

Cannabis substitution is not an established evidence-based treatment for alcohol use disorder, often abbreviated AUD. That is a statement about current treatment evidence, not proof that substitution can never help an individual reduce drinking.

Established options include behavioral treatment and medications selected for the person’s goals and medical circumstances. A 2023 systematic review supports oral naltrexone and acamprosate as first-line medications used with psychosocial care. Disulfiram is another approved option for selected patients, but its evidence, mechanism, and suitability differ; listing the three together should not imply equal effectiveness or interchangeability (McPheeters et al., 2023).

Researchers have also explored cannabis substitution within managed alcohol programs, where alcohol is provided under supervision to reduce harm. The cited study examined need and feasibility, not treatment effectiveness. It does not establish that a self-designed substitution plan works (Pauly et al., 2021).

What if someone already has a diagnosed substance use disorder?

A diagnosis of AUD or cannabis use disorder, abbreviated CUD, changes the question. The priority becomes treatment that addresses impaired control, consequences, and the person’s recovery goals. Evidence does not establish an unsupervised switch between substances as adequate care.

Treatment can support abstinence or, when appropriate, reducing drinking. Medication-assisted approaches require clinical assessment and follow-up. Naltrexone is one studied option for reducing alcohol consumption, although no single medication is the best choice for everyone (review of controlled-drinking treatments; review of diagnosis and pharmacotherapy).

Wanting to drink less is a useful starting point. A clinician can help turn that goal into a plan that addresses withdrawal risk, coexisting conditions, medication options, and support. Seeking that help does not require committing in advance to a particular identity or recovery label.

How to make changes in substance use easier to assess

An honest record of current use and clear criteria for change can make it easier to see what is happening. Without them, cannabis use may gradually increase as alcohol use falls. The result can be a shift in intoxication and impairment rather than the hoped-for reduction in harm.

The following considerations are a monitoring framework, not a validated cannabis-substitution treatment protocol. If physical dependence on alcohol is possible, seek medical advice before abruptly stopping: alcohol withdrawal can be dangerous. Mentioning psychedelics here is not a recommendation to use them.

  • Establish a baseline with a clinician. AUDIT-C or the full AUDIT can screen for alcohol-related risk; CUDIT-R screens for cannabis-related problems. PHQ-9 and GAD-7 may help assess depressive or anxiety symptoms. Liver tests such as ALT, AST, and GGT may be appropriate for a separate clinical reason, but they are not mandatory entry tests for a lifestyle experiment. A positive screen calls for assessment, not a self-diagnosis.

  • Define the actual behavior, rather than relying on “California sober.” Record which substances are involved, how frequently they are used, and whether cannabis is smoked, vaped, or eaten. Keeping track of dose, planned limits, and substance-free days can reveal escalation. No particular dose limit makes cannabis safe, and this framework does not prescribe a dosing schedule.

  • Set a time to review progress, without treating it as an evidence-based waiting period. A 60–90-day period is sometimes proposed, but research has not validated it as the duration needed to form a stable substitution habit or detect CUD. Concerning symptoms or rising use deserve attention sooner.

  • Keep a daily record of cannabis frequency, amount, and product, as well as whether alcohol was consumed. Note sleep, mood, work performance, and relationship strain. A regular weekly review may be easier to interpret than relying on memory at the end of several months.

  • Revisit the same relevant measures with a clinician. An increasing CUDIT-R score or continued problematic drinking may indicate that the intended improvement has not occurred. A score change alone does not establish diagnosis, causation, or overall health benefit. Unexpected results are a reason to adjust care, not a personal failure.

Where California sober can become problematic

Occasional cannabis use can become frequent or heavy use that fills the space left by alcohol. A rising CUDIT-R score may flag problems but cannot diagnose CUD on its own. Recording use from the beginning can help identify escalation. There is no validated 30- and 90-day screening schedule for this substitution strategy. The review of heavy use and dependence does not establish those intervals or a timetable for developing CUD (clinical review).

Another concern is an alcohol use disorder remaining untreated. A person may drink less yet still need help with impaired control, withdrawal risk, or alcohol-related consequences. Naltrexone, acamprosate, and other appropriately selected treatments may be worth discussing (medication review). Common US AUDIT-C positive-screen thresholds are 3 or more for women and 4 or more for men, rather than 4 and 5. Thresholds depend on the tool and setting; a positive result prompts further assessment and does not determine which medication to use (US screening resource).

Cannabis can affect attention, memory, and mood. Gradual changes may be attributed to stress, poor sleep, or aging instead (review of acute and residual neurocognitive effects). Recording symptoms before and after a change can help a clinician assess them. A day-0 and day-90 cognitive or mood check is not a validated causal test: different scores do not prove cannabis caused the difference.

The belief that cannabis is healthier than alcohol can also make its own risks easier to dismiss. Frequent or heavy use has acute and potentially longer-lasting cognitive consequences, although their size and persistence vary (review of cognitive effects). Written limits can help someone notice when their use changes; evidence does not establish that setting a maximum dose or frequency alone guarantees harm reduction.

Who might consider the idea, and who faces greater risks?

The approach is sometimes discussed for adults who drink at risky levels but do not meet AUD criteria and want to cut back. It is also proposed as a temporary harm-reduction strategy while someone waits for an addiction appointment.

Neither proposed use has been established by direct long-term treatment evidence. The acute laboratory trial is not a recommendation for people to start or increase cannabis while waiting for care. A primary care clinician or treatment referral service may help with interim support that does not depend on adopting a new substance.

Particular circumstances call for caution or make the approach inappropriate:

  • With diagnosed AUD or CUD, discuss evidence-based treatment with an addiction clinician rather than relying on substitution alone.
  • A history of cannabis-related psychosis, a schizophrenia-spectrum condition, or a close family history warrants particular concern. Cannabis use is associated with psychosis, with stronger associations for earlier and more frequent use (CDC mental health guidance).
  • During pregnancy or when planning pregnancy, avoid alcohol and cannabis and discuss support for stopping with a clinician. Cannabis is not a safe alcohol replacement in pregnancy (CDC pregnancy guidance).
  • If receiving psychiatric or psychological care, discuss cannabis with the treating clinician and prescriber, including possible medication interactions and changes in symptoms.
  • Adolescents and young adults face developmental concerns and should not be encouraged to substitute cannabis for alcohol (review of adolescent evidence). Roughly age 25 is sometimes used in discussions of brain development, but it is not a sharp threshold after which cannabis becomes harmless.
  • Do not drive, operate machinery, or perform safety-critical work while impaired. Choosing a different intoxicating substance does not remove that responsibility.

A different online consumption plan cannot resolve these risks. A conversation about the person’s health and goals is more useful than deciding whether they qualify for a lifestyle label.

What to assess before a change and at follow-up

How you feel matters, but it cannot establish the total benefit of substituting one substance for another. Some proposed plans compare day-0 and day-90 measurements under similar conditions. Standardized information can complement memory and personal impressions, but this exact schedule is not a validated routine program.

Testing should follow the person’s clinical needs. Laboratory results cannot certify that substance use is safe or that a substitution strategy has reduced overall harm. There is no reason to order a large panel simply because it appears in a wellness checklist.

  • Liver tests: ALT, AST, and GGT can help assess liver health when clinically indicated. They do not specifically measure alcohol intake. Abnormal values can have other causes, and normal values do not exclude harmful drinking. A change after reducing alcohol needs to be interpreted alongside the history and other findings.

  • Blood lipids: Triglycerides may be relevant when assessing metabolic health or the effects of heavy drinking. A better lipid result would not establish a cannabis benefit, and there is no characteristic lipid pattern that validates California sober as a strategy.

  • Glucose measurements: HbA1c and fasting glucose assess aspects of blood sugar regulation. They should not be presented as tests expected to improve through cannabis substitution. Observational associations between cannabis and glucose regulation are mixed and cannot establish a beneficial metabolic effect.

  • Morning cortisol: Alcohol can affect the hypothalamic-pituitary-adrenal axis involved in stress regulation, but a cortisol result is not specific to either alcohol or cannabis. Routine morning cortisol testing has no established clinical role in evaluating this substitution approach.

  • Structured screening: AUDIT-C or AUDIT, CUDIT-R, PHQ-9, and GAD-7 address alcohol-related risk, cannabis-related problems, depressive symptoms, and anxiety symptoms. They measure different things from blood tests and can support a clinical conversation. They do not replace a diagnostic assessment.

Improved laboratory or screening results can be encouraging, but they do not prove that all risks have fallen. Worsening scores, unexpected symptoms, or continued impairment deserve discussion rather than another unsupervised experiment. Follow-up is useful when it helps make a better decision about care.

When addiction care matters more than a wellness trend

Possible AUD, CUD, a mental health crisis, or substance use that is harming work or relationships deserves professional assessment. A self-designed plan cannot substitute for that care. Help can begin with primary care, an addiction service, or a referral line.

In the United States, call or text 988 for crisis support. SAMHSA’s National Helpline, 1-800-662-HELP (4357), provides round-the-clock treatment referral and information. These are US services; readers elsewhere should use the appropriate local service (SAMHSA information; current National Helpline page).

Recording use and reviewing changes can support treatment. The most useful outcome is not a particular label or score, but a plan that addresses the person’s actual risks and goals.

Frequently asked questions

What does California sober mean?

It usually means avoiding alcohol and other drugs while continuing to use cannabis. Some definitions include psychedelics. There is no standardized version. People may assume cannabis is less addictive or less harmful than alcohol, but that cannot be treated as a universal comparison. CUD is a recognized condition, and psychedelics have their own risks. California sober is not an established addiction treatment.

Where did the term California sober come from?

Demi Lovato’s public comments brought the term wider attention in 2021; Lovato later moved away from that approach. Research on replacing alcohol with cannabis predates the label’s popularity, including papers published in 2009 and 2015. The term is cultural shorthand, not a clinical category.

Does California sober work?

Some medical cannabis users report replacing alcohol with cannabis, but observational results vary by context (substitution study). A newer randomized laboratory trial found less immediate alcohol consumption after THC than after placebo. That is not evidence of durable recovery or lower long-term harm. Cannabis substitution has not become an established AUD treatment; medication and psychosocial care have a different evidence base (AUD pharmacotherapy review).

How long does it take to notice a change?

Someone may notice a change in drinking within weeks, but there is no established timetable for lasting benefit from substitution. A proposed 60–90-day period is not a validated habit-formation window, nor does it predict when CUD might develop. The clinical review sometimes cited for those claims does not establish either timeline (review of heavy cannabis use and dependence). Seek help for concerning changes when they occur rather than waiting for a planned endpoint.

Who should not try it on their own?

People with AUD or CUD, previous cannabis-related psychosis, or pregnancy should not rely on a self-directed switch. Adolescents and young adults should not be encouraged to use cannabis as an alcohol substitute. Research on cognition in younger users is nuanced: average effects can diminish with abstinence, but that does not establish safety for all individuals or address every outcome, such as psychosis (Scott et al., 2018; Lorenzetti et al., 2020).

What are the risks?

Cannabis-related problems can develop while an underlying AUD remains inadequately treated. Treatments such as naltrexone or acamprosate may be appropriate for some people with AUD (treatment review).

Frequent or prolonged cannabis use is also associated with cognitive problems, including memory difficulties. The magnitude, persistence, and contribution of other factors vary. Research should not be read as showing irreversible cognitive injury in every person who uses cannabis (young-adult cognitive evidence; clinical review).

Is it actually healthier than drinking alcohol?

It depends on what is being compared. Replacing daily heavy drinking with occasional cannabis changes a person’s exposure differently from replacing modest drinking with daily high-potency cannabis. The comparison also depends on age, mental health, other medications, and impairment-related activities. There is no general recommendation to switch, and a favorable change in one risk does not establish improvement in total health.

Which biomarkers are relevant?

A clinician may consider ALT, AST, GGT, lipids, or HbA1c when there is a clinical reason to assess liver or metabolic health. There is no validated California-sober blood-test panel. Normal or improved results cannot show that cannabis use is harmless, and routine morning cortisol testing is not established for this purpose.

Alcohol and cannabis screening, mood and anxiety assessment, and discussion of daily functioning can provide information that blood tests miss. The useful combination depends on the person; it is not necessary to complete every questionnaire and laboratory test to ask for help.

Literature and sources

  • Reiman A (2009). Cannabis as a substitute for alcohol and other drugs. Harm reduction journal, 6, 35. https://doi.org/10.1186/1477-7517-6-35
  • Lau N, Sales P, Averill S, Murphy F, Sato SO, Murphy S (2015). A safer alternative: Cannabis substitution as harm reduction. Drug and alcohol review, 34(6), 654-9. https://doi.org/10.1111/dar.12275
  • Risso C, Boniface S, Subbaraman MS, Englund A (2020). Does cannabis complement or substitute alcohol consumption? A systematic review of human and animal studies. Journal of psychopharmacology (Oxford, England), 34(9), 938-954. https://doi.org/10.1177/0269881120919970
  • Lucas P, Walsh Z, Crosby K, Callaway R, Belle-Isle L, Kay R, Capler R, Holtzman S (2016). Substituting cannabis for prescription drugs, alcohol and other substances among medical cannabis patients: The impact of contextual factors. Drug and alcohol review, 35(3), 326-33. https://doi.org/10.1111/dar.12323
  • Lucas P, Baron EP, Jikomes N (2019). Medical cannabis patterns of use and substitution for opioids & other pharmaceutical drugs, alcohol, tobacco, and illicit substances; results from a cross-sectional survey of authorized patients. Harm reduction journal, 16(1), 9. https://doi.org/10.1186/s12954-019-0278-6
  • Gunn RL, Aston ER, Metrik J (2022). Patterns of Cannabis and Alcohol Co-Use: Substitution Versus Complementary Effects. Alcohol research : current reviews, 42(1), 04. https://doi.org/10.35946/arcr.v42.1.04
  • Scott JC, Slomiak ST, Jones JD, Rosen AFG, Moore TM, Gur RC (2018). Association of Cannabis With Cognitive Functioning in Adolescents and Young Adults: A Systematic Review and Meta-analysis. JAMA psychiatry, 75(6), 585-595. https://doi.org/10.1001/jamapsychiatry.2018.0335
  • Lovell ME, Akhurst J, Padgett C, Garry MI, Matthews A (2020). Cognitive outcomes associated with long-term, regular, recreational cannabis use in adults: A meta-analysis. Experimental and clinical psychopharmacology, 28(4), 471-494. https://doi.org/10.1037/pha0000326
  • Kuhns L, Kroon E, Colyer-Patel K, Cousijn J (2022). Associations between cannabis use, cannabis use disorder, and mood disorders: longitudinal, genetic, and neurocognitive evidence. Psychopharmacology, 239(5), 1231-1249. https://doi.org/10.1007/s00213-021-06001-8
  • Kroon E, Kuhns L, Hoch E, Cousijn J (2020). Heavy cannabis use, dependence and the brain: a clinical perspective. Addiction (Abingdon, England), 115(3), 559-572. https://doi.org/10.1111/add.14776
  • Meier MH, Caspi A, R Knodt A, Hall W, Ambler A, Harrington H, Hogan S, M Houts R, Poulton R, Ramrakha S, Hariri AR, Moffitt TE (2022). Long-Term Cannabis Use and Cognitive Reserves and Hippocampal Volume in Midlife. The American journal of psychiatry, 179(5), 362-374. https://doi.org/10.1176/appi.ajp.2021.21060664
  • McPheeters M, O’Connor EA, Riley S, Kennedy SM, Voisin C, Kuznacic K, Coffey CP, Edlund MD, Bobashev G, Jonas DE (2023). Pharmacotherapy for Alcohol Use Disorder: A Systematic Review and Meta-Analysis. JAMA, 330(17), 1653-1665. https://doi.org/10.1001/jama.2023.19761
  • Pauly B, Brown M, Chow C, Wettlaufer A, East Side Illicit Drinkers Group for Education (EIDGE), Graham B, Urbanoski K, Callaghan R, Rose C, Jordan M, Stockwell T, Thomas G, Sutherland C (2021). ‘If I knew I could get that every hour instead of alcohol, I would take the cannabis’: need and feasibility of cannabis substitution implementation in Canadian managed alcohol programs. Harm reduction journal, 18(1), 65. https://doi.org/10.1186/s12954-021-00512-5
  • Kranzler HR, Soyka M (2018). Diagnosis and Pharmacotherapy of Alcohol Use Disorder: A Review. JAMA, 320(8), 815-824. https://doi.org/10.1001/jama.2018.11406
  • Palpacuer C, Duprez R, Huneau A, Locher C, Boussageon R, Laviolle B, Naudet F (2018). Pharmacologically controlled drinking in the treatment of alcohol dependence or alcohol use disorders: a systematic review with direct and network meta-analyses on nalmefene, naltrexone, acamprosate, baclofen and topiramate. Addiction (Abingdon, England), 113(2), 220-237. https://doi.org/10.1111/add.13974
  • Dellazizzo L, Potvin S, Giguère S, Dumais A (2022). Evidence on the acute and residual neurocognitive effects of cannabis use in adolescents and adults: a systematic meta-review of meta-analyses. Addiction (Abingdon, England), 117(7), 1857-1870. https://doi.org/10.1111/add.15764
  • Kroon E, Kuhns L, Cousijn J (2021). The short-term and long-term effects of cannabis on cognition: recent advances in the field. Current opinion in psychology, 38, 49-55. https://doi.org/10.1016/j.copsyc.2020.07.005
  • Lorenzetti V, Hoch E, Hall W (2020). Adolescent cannabis use, cognition, brain health and educational outcomes: A review of the evidence. European neuropsychopharmacology : the journal of the European College of Neuropsychopharmacology, 36, 169-180. https://doi.org/10.1016/j.euroneuro.2020.03.012
  • Substance Abuse and Mental Health Services Administration. (n.d.). National Helpline for mental health, drug, alcohol issues. https://samhsa.gov/find-help/national-helpline
  • Metrik J, Aston ER, Gunn RL, Swift R, MacKillop J, Kahler CW (2026). Acute Effects of Cannabis on Alcohol Craving and Consumption: A Randomized Controlled Crossover Trial. American Journal of Psychiatry, 183(2), 134–143. https://doi.org/10.1176/appi.ajp.20250115 Publication

Additional evidence and support links:

Further support