Aim – To describe the rationale for introducing molecular genetic analyses to psychiatric epidemiology. This offers new possibilities for aetiological research on common mental disorders. Method – In addition to the traditional variables used in field surveys, it is now possible to include molecular genetic information. This is currently done by looking for allelic associations rather than by linkage analysis, and can be directed either at traits conferring susceptibility or at states. The purpose is to identify quantitative trait loci (QTLs) by examining candidate genes. An alternative strategy is genome scanning, which can identify genes by their chromosomal position with increasing resolution. Results – Some associations have already been reported in the literature, linking personality traits with particular alleles: extraversion or novelty-seeking with a polymorphism of the dopamine DRD4 receptor gene; and neuroticism with a polymorphism of the serotonin transporter gene. Conclusions – These findings are of major interest, but cannot yet be looked upon as confirmed. What is significant for epidemiology is the opportunity to link behavioural and psychiatric variables with genes influencing biochemical and physiological processes in the brain, and to do so at the population level. For further allelic association studies, there are four principal requirements: valid measures of phenotypes; replication of findings across diverse populations; more candidate genes; and exploration of the interaction between genotype and environmental exposures from conception to late life. Through such research, psychiatric epidemiology can now investigate biopsychosocial phenomena.