Scientists who work with toxic substances, as well as people who keep (or whose family members keep) venomous animals are (or should be) familiar with the LD50 designation, originally called the DL50 system for dosis letalis 50%.
The LD50 of a toxin is the dose at which 50% of the test subjects that ingest the substance die. Invented by J.W. Trevan, a British pharmacologist who published his research on lethal doses in 1927, the LD50 system expresses toxicity in terms of mass of the toxin per mass of the test subject, usually in milligrams (mg) of the substance per kilogram (kg) of body mass.
Because the LD50 is a median, some study subjects will survive a higher dose, while other subjects will be killed by doses lower than the median. The LD50 can also be influenced by the species being studied and the way a toxin is administered.
The LD50 system has generated considerable controversy, as calculation of an LD50 traditionally involved animal testing. An alternative to traditional animal testing is using very early stage chick embryos to measure lethal dosage.
Another alternative might be growing cells or tissues in bottles or Petri dishes filled with a medium containing essential nutrients. To test toxic dosages, known quantities of a substance could be infused into the cell or tissue culture, and the death of cells would be measured by molecular or cellular means.
This method might not be effective for testing neurotoxins, however, because neurotoxins act on complex neural systems not necessarily present in cell or tissue culture.
For an interesting chart on the LD50 values of a number of venoms, look here.