A genetic contribution to schizophrenia is the best established aetiological clue for this common and serious condition. Purely statistical attempts to establish the mode of inheritance have yielded inconclusive results although SML models where a major gene is the sole cause of concordance between relatives can be effectively excluded. The question of genetic heterogeneity is also unresolved. Genetic marker studies offer potentially productive strategies for detecting major genes for schizophrenia or schizophrenia subtypes, but also offer the possibility of detecting contributory (minor) susceptibility loci. The most consistent genetic marker finding, to date, of an association between HLA A9 and paranoid schizophrenia may fall into the latter category. This and other possible associations discovered by recent population studies (e.g. with complement factors) merit further investigation. On the other hand, the findings of linkage studies, including those with HLA, have been predominantly negative. Recent advances in recombinant DNA technology, and the potential availability of many more polymorphisms, make genetic marker studies an increasingly attractive prospect in schizophrenia. However, as in other disorders with complex and non-Mendelian patterns of inheritance, there remain formidable problems in the statistical analysis of the results.