Recent clinical reports on cinchophen toxicosis leave the impression that injury to the liver is an established and characteristic phenomenon of the drug's action. Unfortunately, however, much of the writing on this phenomenon is distinguished by the post hoc ergo propter hoc type of reasoning. While such reasoning errs on the side of safety in medication, it has little or no weight as scientific evidence. Conjectural and deductive reasoning cannot take the place of the controlled experiment. The injury to the liver should be experimentally reproducible if the effect is to be ascribed to a direct action of cinchophen. That is, a direct relationship of cause and effect should be demonstrated. But this has not been achieved, even with doses of the drug which would be therapeutically undesirable or prohibitive, as this paper will show. However, deleterious effects on young growing animals on a dietary containing large amounts of the