"Don't do drugs" treats every substance as interchangeable — equally dangerous, equally worth an identical blanket warning. It isn't true, and it isn't even close. Some legal drugs cause more measurable harm than some illegal ones. Some drugs classified alongside heroin have never caused a documented overdose death. This piece exists to replace the flat warning with the actual picture, using the most rigorous published attempt to measure and compare drug harm directly.
The most rigorous attempt to actually measure this
In 2010, a panel of drug-policy and addiction experts led by British psychiatrist and neuropsychopharmacologist David Nutt published a landmark study in The Lancet, scoring 20 drugs across 16 separate criteria — nine measuring harm to the user themselves (addiction potential, mortality risk, damage to health) and seven measuring harm to others (crime, family and community damage, economic cost). Each drug received a combined score out of 100. It remains, over a decade later, the most methodologically serious public attempt to rank drug harm on evidence rather than legal status or public perception.
Alcohol scored highest overall — more harmful, by this combined measure, than heroin or crack cocaine — driven mainly by its harm to others (drunk driving deaths, violence, family and economic damage): about 46 of its 72 points came from harm to others. Its harm to drinkers themselves (about 26 points) was also substantial, though below heroin, crack cocaine and methamphetamine on that measure alone. Tobacco outscored cannabis. Several drugs most people associate with maximum danger and severe legal punishment — LSD, psilocybin mushrooms, MDMA — landed near the very bottom of the entire list.
Legal status and actual harm turned out to be only loosely related. That gap is the single most important thing to understand before reading anything else in this category.
Test yourself before you see the real numbers
The most convincing way to feel this gap is to notice your own assumptions first. Eight head-to-head match-ups, drawn directly from the study's published overall harm scores — pick the one you think scored higher, then see the real number.
The full ladder, for the drugs covered in this category
Not every substance in this category was part of the original 2010 study — DMT, 5-MeO-DMT, ayahuasca, raw opium, and fentanyl specifically weren't among the 20 drugs scored, mostly because population-level harm data for them is thinner or, for fentanyl, because its current overdose crisis mostly postdates the study. The bars below show the real published overall scores (out of 100) for the substances in this category that were included.
Higher score = more total measured harm in this specific 2010 UK analysis. This is one rigorous study, not an eternal law of nature — dosage, frequency, method of use, drug purity, and individual health all change real-world risk enormously, and the drug supply itself has shifted since 2010 (fentanyl contamination, in particular, has made the opioid landscape considerably more dangerous than these numbers alone would suggest).
What this ladder doesn't tell you
A low score on this scale is not a safety certificate. It means "lower than the other 19 or so drugs studied here, on these particular criteria, in this particular country, at this particular time" — not "safe." Psilocybin scored low here partly because it has an extremely low overdose and addiction potential, which is genuinely true and worth knowing — but it can still trigger a severe psychological crisis in the wrong setting or the wrong person, covered honestly in this category's psilocybin piece. Context — dose, setting, individual health, drug interactions, and what's actually in the substance you're taking — changes real risk more than almost anything else, for every drug on this list.
The most rigorous published attempt to score drug harm found that legal status is a poor predictor of actual danger: alcohol topped the list, tobacco beat cannabis, and several severely punished drugs (LSD, psilocybin, MDMA) scored near the bottom. This doesn't mean any drug is "safe" — it means the categories most people use to think about drug risk (legal vs. illegal) don't track the actual evidence very well, and every piece in this category tries to replace that broad category with the real, specific picture for that one substance.
Try this: before reading any individual piece in this category, notice which substances you already assumed were "worse" or "safer" than others — then check that assumption against what the piece actually says.
Sources. Harm scores and methodology reflect Nutt DJ, King LA, Phillips LD, "Drug harms in the UK: a multicriteria decision analysis," The Lancet, 2010 (376: 1558–1565), conducted by the Independent Scientific Committee on Drugs.
- 🌐 Search "Nutt 2010 Lancet drug harms multicriteria" for the full published study and complete 20-drug table
- 🌐 DrugScience.org.uk — the ongoing independent research organization founded by David Nutt
- 🌐 See each individual piece in this category for substance-specific history, dangers, and (where relevant) therapeutic research