In the mid-1980s, it was widely predicted that family, twin and adoption studies of schizophrenia would be rendered obsolete by the application of molecular genetic techniques to this condition. However, the anticipated advances have yet to occur. Consequently, there has been a revival of interest in the classical genetic epidemiological approach to schizophrenia. Larger and better-designed investigations have been carried out, and new technologies such as brain imaging and neurophysiology have been incorporated into some of these. Family, twin and adoption studies have attempted to answer the following questions:

1 Do genetic factors contribute to the aetiology of schizophrenia? If so,

2 What is the relative contribution of genetic and environmental factors?

3 What is the mode of inheritance?

4 What exactly is inherited?

Investigating whether a genetic effect is present

Since the studies that have addressed the first question have often been reviewed, we will only discuss the major findings and the main sources of bias.

Family studies

Family studies (reviewed by Gottesman and Shields, 1982; Kendler and Tsuang, 1988; Gottesman, 1991; Kendler and Diehl, 1993; McGuffin et al., 1994a, 1995) generally depend on the calculation of the lifetime expectation or morbid risk of schizophrenia in the relatives of probands. The lifetime morbid risk is basically the number of affected relatives divided by the total number of relatives, with an adjustment made for the fact that not all relatives will have passed through the period of risk at the time that they are studied.