Home > Cannabis & Mental Health > Marijuana vs Alcohol
Marijuana vs Alcohol: Which Is Actually Safer?
A Head-to-Head Comparison Using Peer-Reviewed Data
By Dr. Ryan S. Sultan, Assistant Professor of Clinical Psychiatry
Columbia University Irving Medical Center
NIH NIDA-Funded Cannabis Researcher | Published in JAMA Network Open & Pediatrics
|
Neither is "safe." They cause different harms. Alcohol kills more people (95,000/year in the US) and causes more violence and organ damage. Marijuana carries higher risk for psychosis, is more damaging to the developing brain, and 10.4% of young adults now use it daily — triple the daily alcohol rate. The "which is safer" framing is misleading; the real question is what specific harms each one causes. |
Contents:
Why the Question Is Wrong | Head-to-Head Comparison | Addiction | Overdose & Death | Brain Damage | Mental Health | Driving | Physical Health | For Teenagers | The Daily Use Problem | FAQ
Why "Which Is Safer?" Is the Wrong Question
This is the most common question I get from patients, parents, and journalists. It is also the most misleading question in substance use policy.
Asking "is weed safer than alcohol?" is like asking "is a car crash safer than a fall from a building?" The answer depends on the speed of the car, the height of the building, and who is inside. Both substances cause real harm. They cause different types of harm. And the harm depends enormously on who is using, how much, how often, and at what age.
What I can do is lay out the evidence for each domain of harm — side by side — so you can make an informed comparison. That is what this article does.
The Head-to-Head Comparison
| Domain | Alcohol | Marijuana |
|---|---|---|
| Annual US deaths | 95,000 (CDC, 2020). Third leading preventable cause of death | Difficult to quantify; no confirmed lethal overdose deaths. Some driving fatalities |
| Lethal overdose? | Yes. ~2,200 alcohol poisoning deaths/year. Suppresses brainstem respiration | Effectively no. No confirmed cases of fatal THC overdose alone. LD50 is impractically high |
| Addiction rate | ~15% of all users develop alcohol use disorder | ~9-10% of all users; ~30% of regular users develop CUD |
| Physical withdrawal danger | Can be fatal. Delirium tremens, seizures, death without medical supervision | Not dangerous. Uncomfortable (insomnia, irritability, anxiety) but not life-threatening |
| Psychosis risk | Alcohol-induced psychosis exists but is less common | Higher risk. 2.19x hazard for psychotic disorders; 50% conversion to chronic condition |
| Adolescent brain damage | Damages developing brain, particularly cerebellum and white matter | May be worse for teens. 8 IQ point loss, hippocampus shrinkage, disrupts the endocannabinoid construction system |
| Violence | Major driver. Involved in ~40% of violent crimes, 37% of sexual assaults | Not typically associated with violence (may reduce aggression acutely) |
| Liver damage | Yes. Cirrhosis, hepatitis, liver failure — ~25,000 deaths/year | No established liver toxicity from cannabis |
| Cancer | Yes. Established carcinogen (liver, breast, colon, esophageal, head/neck) | Smoked cannabis may increase lung cancer risk; evidence less conclusive |
| Driving impairment | Severe. ~10,500 deaths/year. Well-established dose-response | Moderate. Increases crash risk ~1.3-2x. Harder to measure; users may compensate |
| Depression/anxiety | Worsens both with chronic use; depressant effect | Worsens both with chronic use; nondisordered use: aOR 1.86 for depression (Sultan 2023) |
| Daily use rate (young adults) | 3.4% (Patrick et al., Addiction, 2025) | 10.4% — triple the daily alcohol rate |
Addiction: Head to Head
Alcohol has a higher overall addiction rate:
| Metric | Alcohol | Cannabis |
|---|---|---|
| Lifetime addiction rate (all who try) | ~15% | ~9-10% |
| Addiction rate if start as teenager | ~15-20% | ~17% (1 in 6) |
| Addiction rate with daily use | ~30-40% | 25-50% |
| US adults with the disorder | ~29 million (AUD) | ~14 million (CUD) |
| Physical withdrawal severity | Potentially fatal (seizures, delirium tremens) | Uncomfortable, not dangerous |
| Psychological withdrawal | Severe (anxiety, insomnia, cravings) | Moderate-severe (anxiety, insomnia, irritability, cravings) |
Key nuance: While alcohol's overall addiction rate is higher, the daily cannabis use rate among young adults has now surpassed daily alcohol use by 3:1. Patrick et al. (2025, Addiction) found that 10.4% of 19-30 year-olds use cannabis daily, compared to 3.4% who drink daily. More daily users means more people at the highest risk tier for developing dependence.
For more on cannabis addiction specifically: Is Cannabis Addictive? What the Research Shows
Overdose and Death
This is where alcohol is unambiguously more dangerous.
Alcohol:
- 95,000 Americans die from alcohol-related causes annually (CDC) — making it the third leading preventable cause of death
- ~2,200 die from alcohol poisoning alone (acute overdose)
- ~10,500 die in alcohol-impaired driving crashes
- ~25,000 die from alcoholic liver disease
- Alcohol is involved in approximately 40% of violent crime
Cannabis:
- No confirmed cases of fatal overdose from THC alone in the medical literature
- The lethal dose (LD50) for THC is estimated at 15-70 grams — impractically high to consume
- However, cannabis is not harmless: emergency department visits for cannabis have increased significantly, driven by high-potency products, edible overconsumption, and cannabinoid hyperemesis syndrome
- Cannabis-impaired driving does contribute to traffic fatalities, though attributing causation is methodologically difficult
The bottom line: You can die from drinking too much in a single night. You effectively cannot die from consuming too much cannabis in a single night. This is a meaningful and important distinction. But it does not make cannabis "safe" — it makes it less acutely lethal.
Brain Damage: Different Targets
Both substances damage the brain, but they attack different structures through different mechanisms:
| Brain Effect | Alcohol | Cannabis |
|---|---|---|
| Primary mechanism | Direct neurotoxicity (kills neurons); disrupts GABA/glutamate balance | Endocannabinoid system disruption; CB1 receptor downregulation |
| Hippocampus (memory) | Shrinks with chronic use; largely recovers with abstinence | Shrinks with chronic use; may not recover in adolescents |
| Prefrontal cortex | Impaired; largely recovers | Architecture compromised; permanent in adolescent-onset users |
| Cerebellum (coordination) | Major target — cerebellar degeneration common | Minimal direct cerebellar damage |
| White matter | Damaged; largely recovers with years of abstinence | Damaged or mis-laid during development; may not recover in teens |
| IQ decline | Chronic heavy drinking associated with cognitive decline | 8 IQ points lost in persistent adolescent-onset users (Meier 2012); quitting did not restore |
| Recovery potential (adults) | Substantial recovery over months-years of abstinence | Substantial recovery over weeks-months of abstinence |
| Recovery potential (adolescents) | Partial recovery | Limited recovery — drug was present during construction |
The key distinction for teenagers: The developing brain uses its own endocannabinoid system to guide construction. Cannabis hijacks that system. Alcohol damages the brain from the outside; cannabis corrupts the blueprint from within. This is why adolescent cannabis use may produce more lasting cognitive harm than adolescent alcohol use — though both are damaging.
For the full science: Cannabis and the Teenage Brain
Mental Health Effects
Psychosis
Cannabis wins this comparison — and not in a good way. Cannabis carries a substantially higher psychosis risk than alcohol:
- Young-Wolff 2026 (N = 463,396): Cannabis use disorder associated with 2.19x hazard of psychotic disorders
- ~50% of cannabis-induced psychosis converts to a permanent psychotic condition
- Daily high-potency cannabis use: 5x risk of first psychotic episode (Di Forti 2019)
- A Danish national registry study found eliminating CUD could prevent 30% of schizophrenia diagnoses in young men
Alcohol-induced psychosis exists but is less common and more closely tied to withdrawal states and Wernicke-Korsakoff syndrome.
Depression and Anxiety
Both substances worsen depression and anxiety with chronic use:
- Sultan 2023 (N = 68,263): Even nondisordered cannabis use in adolescents — 1.86x higher depression, 2.08x higher suicidal ideation
- Alcohol is a CNS depressant that worsens depression biochemically; involved in ~30% of suicides
- Cannabis disrupts sleep architecture (suppresses REM), serotonin signaling, and endocannabinoid tone — all contributing to depression
Violence and Aggression
Alcohol is dramatically worse:
- Involved in approximately 40% of violent crimes
- Involved in approximately 37% of sexual assaults
- Alcohol disinhibits aggression; cannabis generally does not
- Cannabis acutely reduces aggression in most users (sedation), though chronic heavy use and withdrawal can produce irritability
Driving Impairment
| Metric | Alcohol | Cannabis |
|---|---|---|
| Crash risk increase | 6-15x at legal limit (0.08 BAC); exponential above | 1.3-2x (meta-analyses vary) |
| Annual US driving deaths | ~10,500 | Difficult to isolate; often co-occurs with alcohol |
| Impairment type | Slowed reaction time, impaired judgment, overconfidence | Slowed reaction time, divided attention deficits; users often compensate (drive slower, increase following distance) |
| Detection | Reliable breathalyzer; 0.08 BAC legal limit | No reliable roadside test; THC in blood does not correlate well with impairment |
| Combined with the other | Dramatically worse than either alone. Alcohol + cannabis together produces multiplicative impairment | |
Alcohol is a more dangerous driving impairment by every measure. However, cannabis-impaired driving is not safe — and the combination of both substances is far more dangerous than either alone.
Physical Health Effects
| System | Alcohol | Cannabis |
|---|---|---|
| Liver | Cirrhosis, hepatitis, liver failure. ~25,000 deaths/year | No established liver toxicity |
| Heart | Cardiomyopathy, arrhythmias, hypertension | Increases heart rate 20-50 bpm; may trigger MI in vulnerable individuals |
| Cancer | Established carcinogen (7+ cancer types: liver, breast, colon, esophageal, oral, pharyngeal, laryngeal) | Smoked cannabis: possible lung cancer risk; evidence inconclusive. No established non-pulmonary cancer link |
| Lungs | Not directly toxic to lungs | Smoked cannabis causes chronic bronchitis, cough. Vaping: emerging lung injury data (EVALI) |
| GI system | Gastritis, pancreatitis, GI bleeding | Cannabinoid hyperemesis syndrome (cyclic vomiting) — increasingly common with high-potency use |
| Immune system | Immunosuppression with heavy use | Some immunomodulatory effects; clinical significance unclear |
| Pregnancy | Fetal alcohol spectrum disorders; no safe amount established | Associated with low birth weight and preterm birth; endocannabinoid system active in fetal development |
Alcohol causes dramatically more physical disease. This is not close. Chronic heavy drinking destroys the liver, heart, GI tract, and immune system in ways cannabis does not. However, cannabis is not physically benign — especially smoked cannabis and the emerging problem of cannabinoid hyperemesis syndrome.
For Teenagers: A Special Comparison
The "which is safer" question takes on particular urgency when it comes to adolescents. Many parents reason: "If my teen is going to use something, at least marijuana is safer than alcohol."
This reasoning is flawed. For the developing brain, cannabis may actually be more harmful than alcohol in specific ways:
| Adolescent Harm | Alcohol | Cannabis |
|---|---|---|
| IQ decline | Associated with heavy use, partially reversible | 8 IQ points lost, NOT reversible (Meier 2012) |
| Psychosis risk | Lower | 2.19x higher for psychotic disorders; strongest at ages 13-15 |
| Depression | Worsens depression | 1.86x higher even without addiction (Sultan 2023) |
| Brain development disruption | Damages from outside | Corrupts the endocannabinoid construction system from within |
| Daily use rate (ages 19-30) | 3.4% | 10.4% — 3x higher |
| Acute lethality | Yes — alcohol poisoning kills teens | Extremely rare |
| Sexual assault involvement | Major factor | Not a significant factor |
The honest answer for parents: neither substance is acceptable for a developing brain. If forced to choose a "lesser evil," the answer depends on the specific harm you are most concerned about. Acute danger and violence? Alcohol is worse. Long-term brain development and psychosis? Cannabis may be worse.
The right answer is to delay both until after 25 if possible, and to prioritize reduction in frequency and quantity if use is already occurring.
The Daily Use Problem: Cannabis Is Winning — and That's Not Good
A critical and underappreciated finding from Patrick et al. (2025, Addiction):
|
10.4% of young adults (ages 19-30) now use cannabis daily or near-daily — triple the rate of daily alcohol use (3.4%). This is a historic first. Cannabis has overtaken alcohol as the substance most commonly used on a daily basis by young Americans. |
This matters because the harms of both substances are dose-dependent. Occasional use carries different risk than daily use. And at the daily use level:
- Daily cannabis users have a 25-50% lifetime risk of developing cannabis use disorder
- Daily high-potency cannabis use is associated with 5x the psychosis risk
- My research (Sultan 2026, Pediatrics, N = 162,532) shows a clear dose-response gradient: the more frequently adolescents use, the worse every outcome measured
The "safer than alcohol" narrative may be contributing to this trend. If young people believe cannabis is harmless, they use more of it, more often. The population-level consequence of widespread daily cannabis use may eventually rival the population-level harm of alcohol — not because cannabis is more dangerous per use, but because so many more people are using it so much more often.
Frequently Asked Questions
Is weed safer than alcohol? Just give me a straight answer.
Cannabis is less acutely lethal, less physically destructive, and less associated with violence than alcohol. Alcohol is less likely to cause psychosis, less damaging to the adolescent brain's construction process, and has more established treatment protocols. If "safer" means "less likely to kill you tonight," cannabis is safer. If "safer" means "less likely to permanently alter your brain development," the answer is more complicated — especially for teenagers.
Are edibles safer than drinking?
Edibles avoid the lung damage of smoked cannabis and the liver damage of alcohol, but they carry unique risks: unpredictable onset (30-90 minutes) leads to overconsumption, effects last 6-8 hours making impaired driving more likely over a longer window, and high-dose edibles can trigger severe panic or psychotic episodes. "Safer" in one dimension, riskier in others.
My state legalized marijuana. Doesn't that mean it's safe?
Legal does not mean safe. Alcohol is legal. Tobacco is legal. Both kill hundreds of thousands of Americans annually. Legalization is a policy decision about criminal justice, taxation, and regulation — not a medical endorsement of safety.
I use marijuana instead of drinking. Is that a good trade?
If you were a heavy drinker who switched to occasional cannabis use, you likely reduced your health risk — particularly for liver disease, violence, and acute overdose. If you traded occasional drinking for daily high-potency cannabis use, you may have traded one set of risks for another. The key variables are frequency, quantity, and potency, not just which substance.
What about CBD vs alcohol?
CBD (without THC) does not produce intoxication, addiction, or the mental health risks associated with THC. Comparing CBD to alcohol is misleading because they serve fundamentally different purposes. CBD is not a recreational intoxicant. If the question is "should I use CBD oil instead of drinking wine to relax," that is a reasonable consideration — but CBD products are unregulated, and evidence for CBD's anxiolytic effects is still limited.
About This Article
Written by Dr. Ryan Sultan, a board-certified psychiatrist and NIH NIDA-funded researcher at Columbia University. Dr. Sultan's cannabis research has been published in JAMA Network Open and Pediatrics.
Related articles: Is Cannabis Addictive? | Cannabis & the Teenage Brain | Cannabis & Psychosis | Cannabis & Anxiety | THC Potency: Why Today's Weed Is Different | Cannabis & Mental Health Guide
Concerned About Substance Use?
Dr. Sultan provides expert psychiatric evaluation for cannabis and alcohol-related concerns at Integrative Psych in NYC.
Request Consultation →