The genetic contribution to the etiologies of schizophrenia and bipolar affective disorder (BPAD) has been considered for many decades, with twin, family, and adoption studies indicating consistently that the familial clustering of affected individuals is accounted for mainly by genetic factors. Despite the strong evidence for a genetic component, very little is understood about the underlying genetic and molecular mechanisms for schizophrenia and BPAD. In the early 1990s, after the discovery of "dynamic mutation" or "unstable DNA" as a molecular basis for the genetic anticipation observed in Huntington's disease, myotonic dystrophy, and many others, and the recently rediscovered, albeit still controversial, evidence for genetic anticipation in major psychoses, the genetic epidemiology of schizophrenia and BPAD was re-evaluated to demonstrate strong endorsement for the unstable DNA model. Many of the non-Mendelian genetic features of schizophrenia and BPAD could be explained by the behaviour of unstable DNA, and several molecular genetic approaches became available for testing the unstable DNA hypothesis. However, despite promising findings in the mid-1990s, no trinucleotide repeat expansion has yet been identified as a cause of idiopathic schizophrenia or BPAD.