Level of dosage for a substance, controlled or otherwise, which is lethal for 50% of subjects who take that amount. Data like this is (obviously) usually only available for non-humanoids.

For example, the LD50 of LSD is 46 mg/kg intravenously in mice -- compare to an active dosage of .001 mg/kg (1 microgram/kg) in humans.

There is a wonderful book called Molecules at an Exhibition in which the author does interesting profiles of compounds that appear in everyday things or situations, as well as unusual compounds like sarin, an extremely potent neurotoxin. As I was reading it, I noticed that he would mention the LD50 of almost every one, in sort of a morbid kind of way. I mentioned it to my cousin, who is doing a PhD in biophysics, and he said that it's one of the first things that chemists tend to ask about a compound, as in: "Cool, what's the LD50?"

OBTW, a lethal dose of caffeine is about 10 grams. That's around 100-200 cups of coffee, depending on your body mass, and how strong you like your java.

A measure of the acute toxicity of chemicals, developed in 1927; LD stands for lethal dose. The amount of a substance which when administered to test animals, usually rats or mice, causes the death of half of them. Expressed as weight of the chemical administered per kilogram body weight of the test animal, and animal used and method of administration or exposure:
LD50 (oral, rat) - 5 mg/kg
LD50 (dermal, rabbit) - 10 g/kg.
Useful because different poisons or medicines may have effects of similar intensity but on different tissues or organ systems, so a test is needed that produces an identical effect, namely death.

An LD50 test is performed by administering gradually increasing dosages to successive test populations over a set time frame, usually 14 days, until 50% of the subjects die, after which surviving animals are killed.

The LC50 test is another measure of acute toxicity, but of concentrations in air or water rather than an administered dosage, and is expressed in parts per million (ppm) or milligrams per cubic meter, and the test animal, and the duration of exposure:

LC50 (rat) - 500 ppm/4 hr
LC50 (rabbit) - 5 mg/m3/2 hr.
LD50 is controversial today, for several reasons. The test consumes a large number of animals and to those not inured to laboratory use of animals, seems cruel. It gives little or no information about cause of death. Its reliability is very sensitive to test conditions. Most telling, variations in species-to-species sensitivity call into question the test's relevance for human toxicity. For example, acetaminophen has an LD50 in mice of about 300 mg/kg due to liver necrosis, but in rats the LD50 is about 1000 mg/kg with little liver damage evident.

Nevertheless, use of the LD50 test is widespread not only as a measure of chemical toxicity, but to:

  • classify substances for regulatory purposes including safe transportation and labeling
  • provide information for treatment of acute intoxications
  • standardize certain biological products
  • set dose levels for subsequent toxicity studies
  • provide comparative information on dose response curves
  • provide data for evaluation and validation of alternative test methods.
  • Short for Lethal Dose, an LD 50 value is the amount of a solid or liquid material that it takes to kill 50% of test animals, such as rats or mice, in one dose. LD 50 values are specified in mg/kg along with the subject animal in question and the method of injection.

    For example, the LD 50 value of DDT for rats, injected orally, is 87. So, a dosage of 87 miligrams of DDT for every kilogram the rat weighs has a 50% chance of killing the rat.

    Some facts:

    1. The LD 50 does not scale exactly, either from mice to humans or from 50% to 100%. People vary. People take six times the "lethal dose" (not the LD 50, but a higher dose that in a treating physician's opinion usually causes death) of something and manage somehow to survive because bodies differ.
    2. Human-specific LD 50 data is not available because the FDA won't let drug companies take a pool of 1200 or so humans and feed them a drug until 600 are dead.
    3. Because a single test may kill as many as 1000 animals, the United States and other members of the Organisation for Economic Cooperation and Development agreed in December 2000 to phase out the LD 50 test in favor of alternatives that greatly reduce (or even eliminate) deaths of the test animals.
    4. Even adjusting for the test animal's weight, the LD50 for one species is often quite different from that for another. Thus any LD 50 value gives only a rough estimate of the risk to humans.
    5. The way in which the chemical is administered also has a marked effect on LD 50 values. The chemical may be fed, injected, applied to the animal's skin, etc., and each method usually generates a different LD 50.

    So, even though the LD 50 dosage for rats of a given substance is not a dead-sure (pun intended) way to get yourself killed, never be exposed to an LD 50 dose of a hazardous chemical. Even if you survive you're not going to be in good shape.

    Oh yeah, L.D. 50 is also the snazzy title of Mudvaynes debut album. It ties in neatly with the cover art and the the general theme of the album, which involves such things as the evolution of man and drugs, among other things

    A few LD50 (Lethal dose for 50% of the population studied) values:

  • Alcohol (Ethyl Alcohol): 10.6 g/kg, young rats, orally
  • Antifreeze (ethylene glycol) 6.86 g/kg, rats, orally
  • Table salt (sodium chloride) 3.75 g/kg, rats, orally
  • Marijuana (delta1-3,4-trans-tetrahydrocannabinol) 3.75 g/kg, rats, orally
  • GHB (Gamma-Hydroxybutyrate, as sodium-GHB) 2 g/kg, rats, intraperitoneal
  • Wormwood (thujone, one of the main constituent of absinth) 0.45 g/kg, mice, orally
  • Chocolate (theobromine by itself, although chocolate also contains caffeine) 0.310 g/kg, rats, orally (much higher for cats and dogs)
  • Ketamine (2-(2-Chlorophenyl)-2-(methylamino)cyclohexanone) 0.229 g/kg, rats, orally
  • Ecstasy (MDMA): 0.049 g/kg, rats, intra peritoneal
  • Psychoactive Amanitas (Muscimol): 0.045 g/kg, rats, orally
  • LSD-25 (lysergic acid diethylamide): 0.0165 g/kg, rats, intravenous
  • Strychnos seeds (brucine) 0.001 g/kg, rats, orally
  • Nicotine (3-(1-methyl-2-pyrrolidinyl)pyridine) 0.0003 g/kg, mice, orally
  • Ricin (a high molecular weight protein made from castor beans, it has been used in experimental treatments for cancer and AIDS) 0.000005 g/kg, mice, orally
  • Enya (new age female Celtic bard): 0.0 g/kg (instant death), rats and human alike, aurally

    (compiled from miscellaneous sources, including Erowid and the Registry of Toxic Effects of Chemical Substances (RTECS)...

    A few comments and personal conclusions from this list:

  • You should not drink antifreeze.
  • You should not try to assess the accuracy of these data by trying with your dog/cat/goldfish/neighbourhood rat population/significant other/kids. Unless you are scientifically qualified to conduct this type of experiment.
  • From looking at this list, it should be pretty obvious that LD50 values are to be taken with a huge grain of salt (although no more than 3.75 g/kg if you are an adult-size lab rat): I think we all know that chances to die from a GHB overdose are much higher than from a chocolate overdose, despite appearances of the contrary. The nuance between toxicity and lethality plays a large role in these discrepancies, as well as the essential difference in metabolism between a small rodent (known for thriving amidst major city's most toxic areas) and a much larger biped (also known for thriving in said cities, right next to said toxic areas, but that's not a reason).
  • Despite my best efforts to harmonize the data (trying to pick preferably oral administration on male adult rats), I was not always able to find perfectly matching data due to lack of source or nature of substance involved (rats will eat about anything, but there is a limit).
  • Although LD50 experiments altogether are a much controversial issue, due to its particularly inhuman and cruel aspect, the use of Enya's music should decently be banned from any self-respecting scientist's laboratory.
  • Log in or register to write something here or to contact authors.