There are currently two separate 'transmethylation' hypotheses of schizophrenia. The original hypothesis, dating from 1952, suggested that the disease might be due to the production in the brain of psychotomimetic methylated derivates of catecholamines or indolealkylamines. However, years of intensive search have failed to provide evidence to support this theory. The second hypothesis, more appropriately known as the 'one-carbon cycle hypothesis', dating from 1963, suggests some basic fault in the transmethylation mechanism, i.e. the one-carbon cycle itself. Recent research has produced evidence to support an under-activity of this system in schizophrenia and depression, and a possible over-activity in mania. Transmethylation mechanisms are now becoming important as control systems for the brain: in particular, in coupling receptors to adenylate cyclase (lipid methylation); neurotransmitter release, aspects of chemotactic behavior and dopamine receptors (carboxymethylation); and gene expression (nucleic acid methylation). A basic fault or faults in this system might be expected to have widespread results.