The probability that a teratogen, applied to a pregnant mammal, will produce malformations in the embryo depends on the agent, the dose, the species, the genetic constitution of mother and embryo, and the developmental stage of the embryo. Several drugs (including salicylates and antibiotics) now being used in medical practice are teratogenic in experimental animals, some at doses comparable, on a body-weight basis, to those used therapeutically. Demonstration of teratogenicity in experimental animals can serve as a warning of possible teratogenic effects in man, and as a guide to the types of malformations the drug might produce, but failure to demonstrate teratogenic effects experimentally does not prove the drug's harmlessness to the human embryo. If thalidomide had produced only common malformations, such as cleft lip, its teratogenic nature might still be unrecognized. The final test must be careful follow-up of babies born to mothers who have taken the drug in question.