Alcohol, when measured by total harm to both users and the people around them, consistently ranks as the most damaging substance in multi-criteria analyses that compare it against illegal drugs. A landmark UK study scored alcohol at 72 out of 100 for overall harm, ahead of heroin at 55 and crack cocaine at 54. That finding surprises people who assume legality signals safety, but the science behind it is remarkably consistent across countries and methodologies, and the reasons are worth understanding in detail.
How Researchers Actually Rank Drug Harms
The question “which is worse” sounds simple, but it requires defining what “worse” means. Researchers have developed scoring systems that weigh multiple dimensions of harm: physical damage to the user, likelihood of dependence, mental health effects, injury to others, economic cost, family disruption, and more. The most influential of these studies, published in The Lancet in 2010, used a panel of experts to score 20 substances on 16 criteria grouped into harms to users and harms to others. Heroin, crack cocaine, and methamphetamine did the most damage to individual users. But alcohol dominated the “harm to others” category so heavily that it came out on top overall.1The Lancet. Drug harms in the UK: a multicriteria decision analysis
This was not a quirk of one British panel. A 2023 New Zealand study used the same general approach and reached a similar conclusion: alcohol, methamphetamine, and synthetic cannabinoids topped the harm list for the overall population, with alcohol remaining the most harmful drug even when the analysis separated harm to users from harm to others.2PubMed Central. The New Zealand drug harms ranking study: A multi-criteria decision analysis These rankings do not mean a single dose of alcohol is more dangerous than a single dose of heroin. They mean that when you account for how many people use a substance, how they use it, and how wide the ripple of damage spreads, alcohol causes more total harm than any other drug.
Why Alcohol Dominates Harm-to-Others Scores
The reason alcohol pulls ahead of substances most people consider “harder” is largely about what it does to other people. A heroin user in a locked room is overwhelmingly damaging themselves. A drinker at a bar, behind a wheel, or inside a family home can radiate harm outward in ways that few other substances match.
Violence is a major driver. Alcohol has been linked to violent crimes and domestic violence across many countries, and individuals who are already prone to aggression become considerably more dangerous while drinking.3PubMed Central. Alcohol, Aggression, and Violence: From Public Health to Neuroscience A case-crossover study found a roughly 13-fold increase in the risk of committing criminal violence within 24 hours of drinking, confirming alcohol as a powerful trigger for violent behavior.4PubMed. The role of alcohol and drugs in triggering criminal violence: a case-crossover study Other substances can fuel violence too, particularly stimulants, but the sheer number of people drinking on any given evening means alcohol-related assaults dwarf those linked to other drugs in raw terms.
Traffic safety tells a similar story. Research consistently finds that alcohol increases crash risk and worsens clinical outcomes when crashes do occur.5PubMed. Effects of alcohol and other drugs on driver performance The picture for other substances is more mixed. Cannabis, for instance, has not produced the same unanimous evidence of increased crash risk in epidemiological studies, although combining cannabis with alcohol raises the danger above either substance alone.6PubMed Central. The effect of cannabis compared with alcohol on driving An analysis of fatal crash data found that while being positive for drugs other than alcohol did contribute to crash risk, that contribution was significantly lower than what alcohol produced.7PubMed Central. Drugs and alcohol: their relative crash risk
The Global Toll in Lives and Health
The worldwide burden of alcohol dwarfs the burden of all illicit drugs combined, and the gap is not close. A systematic analysis of data from 195 countries found that in 2016, alcohol accounted for about 99 million disability-adjusted life years lost globally, representing roughly 4% of all health burden worldwide. Drug use, by contrast, accounted for about 32 million. Put another way, alcohol was responsible for about three-quarters of all substance-use-attributable health loss.8PubMed Central. The global burden of disease attributable to alcohol and drug use in 195 countries and territories, 1990–2016
Among young people the ratio narrows, but alcohol still leads. A 2019 analysis found that among people aged 10 to 24, alcohol use was tied to roughly 60,000 deaths and 5.9 million disability-adjusted life years, while drug use contributed about 16,000 deaths and 4.1 million disability-adjusted life years.9PubMed. Global burden attributed to alcohol and drug use among adolescents and young adults, 1990–2019 The mortality gap is roughly four-to-one, though drug-related disability catches up somewhat because the years of healthy life lost per drug death tend to be high in this age group.
Lethality Per Dose Versus Lethality at the Population Level
Here is where the picture gets complicated, and where many arguments about “which is worse” talk past each other. On a per-dose basis, alcohol is actually one of the less acutely lethal substances. The amount of alcohol needed to kill you is hundreds of times your typical recreational dose. Heroin’s lethal dose, by contrast, sits much closer to its recreational dose. A study using the margin-of-exposure approach found that heroin had a benchmark dose of about 2 mg per kilogram of body weight, while alcohol’s was around 531 mg per kilogram. By that measure, heroin is far more dangerous per encounter.10PubMed Central. Comparative risk assessment of alcohol, tobacco, cannabis and other illicit drugs using the margin of exposure approach
Yet when the researchers scaled up from individual risk to population risk, accounting for how many people actually use each substance and at what levels, the picture flipped. Only alcohol fell into the “high risk” category at the population level. Cannabis, at the opposite extreme, had a margin of exposure more than 10,000 times its typical dose, making it far safer on an individual exposure basis than any other substance in the study.10PubMed Central. Comparative risk assessment of alcohol, tobacco, cannabis and other illicit drugs using the margin of exposure approach This distinction between per-dose danger and population-level danger is the single most important nuance in the whole debate. A substance does not need to be especially deadly per use to cause the most total death and disability; it just needs to be used by enough people, frequently enough, in large enough amounts.
Addiction Potential Across Substances
Another dimension of harm is how likely a substance is to hook you once you try it. Data from a large US national survey tracked the lifetime probability of transitioning from first use to dependence for several substances. About 68% of people who ever tried nicotine eventually became dependent. For alcohol, the figure was roughly 23%. Cocaine came in at about 21%, and cannabis at about 9%.11PubMed Central. Probability and predictors of transition from first use to dependence on nicotine, alcohol, cannabis, and cocaine
Alcohol’s capture rate may look modest next to nicotine’s, but context matters. Far more people try alcohol than try cocaine, so a 23% conversion rate applied to hundreds of millions of drinkers produces a staggering absolute number of people with alcohol use disorder. Cocaine’s slightly lower rate applied to its much smaller user base produces far fewer total cases. This is the same pattern seen in the mortality data: per-user risk and population-level impact tell different stories, and public health cares most about the population story.
Organ Damage and Long-Term Physical Health
Chronic heavy drinking damages nearly every major organ system. The liver gets the most attention, but alcohol also injures the heart, pancreas, gut lining, and brain. A comprehensive review described the widespread and significant nature of alcohol-induced tissue and organ injury, emphasizing just how many pathways alcohol uses to cause damage.12PubMed Central. Alcohol abuse: critical pathophysiological processes and contribution to disease burden Alcohol is also a Group 1 carcinogen, linked to cancers of the mouth, throat, esophagus, liver, breast, and colon. Few people associate their evening wine with cancer risk, but the International Agency for Research on Cancer has classified ethanol as carcinogenic for decades.
Other substances cause serious organ damage too, of course. Methamphetamine ravages teeth and cardiovascular tissue. Cocaine stresses the heart. Opioids suppress breathing. But the breadth of alcohol’s damage across organ systems, combined with the duration and frequency of typical use patterns, means it racks up more total chronic disease than most individual illicit drugs. The main exception is tobacco, which produces comparable or greater long-term disease burden through cancer and cardiovascular harm, though it scores lower on the multi-criteria analyses because it causes much less harm to people around the smoker (secondhand smoke aside).
Withdrawal and the Danger of Quitting
One of alcohol’s less-discussed dangers is the withdrawal itself. Abrupt cessation in a heavy drinker can cause seizures, delirium tremens, and death. Opioid withdrawal is famously miserable but rarely fatal on its own. Despite these differences in lethality, the two syndromes share some underlying biology. Both involve marked disruption of the autonomic nervous system, and both respond to the same class of medication (alpha-2 agonists), suggesting a shared pathway of negative reinforcement where the brain adapts to the substance and punishes its removal.13PubMed Central. Similarities in alcohol and opioid withdrawal syndromes suggest common negative reinforcement mechanisms involving the interoceptive antireward pathway Benzodiazepine withdrawal can also be fatal through a similar mechanism, because benzodiazepines act on the same brain receptor system as alcohol. Most other drug withdrawals, including those from stimulants and cannabis, are uncomfortable but not life-threatening.
Effects on the Developing Brain
Adolescent brains appear particularly vulnerable to alcohol. Neuroimaging studies have detected structural abnormalities in brain volume and white matter quality in teens with as little as one to two years of heavy drinking, defined as roughly 20 drinks per month or binge episodes of four to five drinks at a time. Heavy marijuana users in the same age range showed some subtle differences, but generally not the same degree of divergence from non-using peers.14PubMed Central. The influence of substance use on adolescent brain development
A separate imaging study looking at both adults and adolescents reinforced this pattern. Alcohol use was associated with reduced gray matter volume and poorer white matter integrity in adults, with some gray matter effects also visible in adolescents. Cannabis use, by contrast, showed no significant associations with brain structure in either age group.15PubMed Central. Structural neuroimaging correlates of alcohol and cannabis use in adolescents and adults This does not mean cannabis is harmless for young brains, since other research has flagged cognitive and functional effects. But on the specific measure of visible brain structure, alcohol looked worse.
Prenatal exposure adds another dimension. Alcohol can cause a recognized spectrum of physical and cognitive alterations in exposed infants, driven by its disruption of key receptor systems in the developing brain. Cocaine exposure during pregnancy has its own set of consequences, including altered cell development and long-term behavioral changes, plus direct harm to the placenta itself.16PubMed. Toxic effects of prenatal exposure to alcohol, tobacco and other drugs Fetal alcohol spectrum disorder remains one of the most common preventable causes of intellectual disability worldwide, a public health cost almost entirely attributable to a legal substance.
The Economic Picture
Money follows damage. National cost estimates from the US put the annual economic toll of alcohol at about $249 billion, illicit drugs at about $193 billion, and prescription opioids at about $78.5 billion, for a combined substance-use price tag exceeding $740 billion. Alcohol alone accounts for the single largest share.17PubMed. Economic consequences of legal and illegal drugs: The case of social costs in Belgium A Belgian social-cost analysis reached the same structural conclusion from European data: alcohol and tobacco impose higher costs on society than illegal drugs do.17PubMed. Economic consequences of legal and illegal drugs: The case of social costs in Belgium
These figures reflect healthcare expenses, lost productivity, criminal justice involvement, and the less tangible costs of family disruption and child welfare. They also reflect the sheer scale of legal substance use. Illegal drugs, by their nature, reach fewer users. If heroin or methamphetamine were as widely available and socially normalized as alcohol, their cost figures would look very different, though no researcher can say exactly how different.
Where Specific Drugs Are Unambiguously Worse
None of this means every drug comparison favors alcohol as the greater villain. Individual substances excel at specific types of harm in ways alcohol does not. Heroin and fentanyl kill more efficiently per dose than almost anything else, and the ongoing opioid crisis has pushed drug overdose deaths past alcohol-related deaths in some countries and age groups. Methamphetamine causes psychotic episodes more reliably than alcohol does. Cocaine triggers sudden cardiac events in young, otherwise healthy users. Synthetic opioids now contaminate drug supplies unpredictably, adding a layer of danger that has no alcohol equivalent.
The mental health picture is also substance-specific. Cannabis use disorder, particularly at heavy levels, has been associated with elevated risk of psychosis in people already vulnerable to it. One study of high-risk individuals found that those with cannabis abuse or dependence had higher rates of converting to psychosis than non-users, though when alcohol use was added to the statistical model, the cannabis association weakened and became non-significant, suggesting the two substances may interact or confound each other in psychiatric outcomes.18PubMed Central. Alcohol confounds relationship between cannabis misuse and psychosis conversion in a high-risk sample Disentangling the psychiatric effects of substances people often use together remains one of the harder problems in this field.
Polysubstance Use Complicates Every Comparison
Real-world substance use rarely involves just one drug. People who use illicit drugs frequently drink as well, and the combination can be more dangerous than either substance alone. Swedish data on drug-poisoning deaths found that about 78% of cases involved multiple substances. Among those who died from a single substance alone, alcohol was the most common culprit. And a telling pattern emerged: as the number of additional drugs in a person’s blood increased, the blood-alcohol level needed to prove fatal dropped sharply, from about 3.1 g/L in pure alcohol deaths down to about 1.25 g/L when six or more other substances were present.13PubMed Central. Similarities in alcohol and opioid withdrawal syndromes suggest common negative reinforcement mechanisms involving the interoceptive antireward pathway Alcohol acts as a multiplier, lowering the lethal threshold for other substances while those substances do the same to it.
This polysubstance reality means that clean comparisons between “drugs” and “alcohol” are partly artificial. In the emergency room, in the morgue, and in the family, the damage rarely comes from one chemical acting alone.
Why Legality Warps Our Sense of Danger
Cross-national surveys of substance use have highlighted a basic measurement problem: people underreport illegal drug use far more than they underreport drinking. Research using WHO data noted that despite methodological efforts to minimize this bias, differences in willingness to report likely persist and color comparisons between countries.19PLOS Medicine. Toward a Global View of Alcohol, Tobacco, Cannabis, and Cocaine Use: Findings from the WHO World Mental Health Surveys This means drug-related harms are probably somewhat undercounted relative to alcohol harms in epidemiological data, though not enough to close the enormous gap in overall burden.
Legality also shapes the cultural framing of danger in ways that have nothing to do with pharmacology. Alcohol has had millennia to become embedded in human social life. Our ancestors gained the ability to metabolize ethanol about 10 million years ago, when early hominids began spending time on the forest floor and encountered fermenting fruit.20PubMed Central. Hominids adapted to metabolize ethanol long before human-directed fermentation That deep evolutionary familiarity, combined with centuries of cultural integration, makes alcohol feel normal in a way that newer or illegal substances never do. People instinctively calibrate “dangerous” against “illegal” rather than against actual harm data, and the result is a persistent public perception gap where a glass of wine seems fundamentally different in kind from a drug, even though pharmacologically it is one.
Where the Evidence Gets Genuinely Uncertain
Researchers acknowledge several limitations in the harm-ranking approach. The expert panels that score substances can be influenced by cultural assumptions, and different weightings of the criteria can shuffle the rankings. The harms of illegal drugs may be partly artifacts of prohibition itself: contaminated supplies, lack of dosing information, barriers to seeking medical help, and criminal-justice consequences all inflate the harm score of substances that might be somewhat less damaging under a regulated regime. Alcohol’s harm score, conversely, might look different if it were suddenly made illegal and driven underground, where quality control disappears and users avoid hospitals.
The comparison also depends on which drugs you lump together. “Drugs” is not a single category. Cannabis and fentanyl have about as much in common as beer and bleach. Asking whether “drugs” are worse than alcohol is a bit like asking whether “food” is healthier than broccoli. The honest answer is always “it depends on which drug, in what dose, used by whom, in what pattern, in what context.” What the evidence does say, clearly and repeatedly, is that alcohol’s combination of widespread availability, social normalization, capacity for chronic organ damage, and outsized harm to bystanders makes it the largest single contributor to substance-related suffering at the population level, even as individual drugs can be far more lethal per encounter.