The concept of a chemical cause of schizophrenia is as old as the Hippocratic School. The idea gained favour after Thudicum postulated that insanity was caused by poisons fermented within the body and that chemical studies could isolate the poisons and discover antidotes. Research with chemical compounds capable of inducing ‘experimental psychoses’ led to a revival of interest in the toxic theory, and the development of new biochemical techniques intensified the search for a biochemical lesion in schizophrenia. Claims that a variety of abnormal metabolites or unusual amounts of normal metabolites are found in mental patients have been made, including: adrenal corticosteroids, imidazoles, adrenal medullary and other catecholamines, adrenochrome and other possible catecholamine metabolites, 5-hydroxy tryptamine, various indoles, copper-containing compounds (ceruloplasmin) and even histamine. It was also claimed that transfusions of normal blood caused transient improvement of schizophrenics and, conversely, that blood from schizophrenic patients contained a toxic substance which could produce abnormal behaviour in human volunteers, trained monkeys and rats. Toxic effects have been reported following administration of fractions from ‘schizophrenic’ blood or other body fluids in a variety of biochemical and physiological procedures and in a variety of organisms; mice, tadpoles, crabs, pigeons, spiders, larvae, seeds, molds and fungi. Most of these claims have been disputed: the abnormal substances are not present, or not present in higher concentrations in schizophrenics, or not capable of inducing measurable behavioural changes. Appealing as the chemical concept is, no unequivocal, substantiated evidence for it has yet been produced.