We thank Professors Craddock and Owen for the insightful comments on the possible molecular genetic basis of the relation between epilepsy and psychosis. In most clinical studies of psychosis in patients with epilepsy, individual psychotic vulnerabilities are rarely concerned compared with epilepsy-related factors. However, several large studies have recently demonstrated genetic vulnerabilities to psychosis even in patients with epilepsy. ^{Reference Adachi, Matsuura, Hara, Oana, Okubo and Kato1,Reference Qin, Xu, Laursen, Vestergaard and Moriensen2} Our recent work ^{Reference Adachi, Akanuma, Ito, Kato, Hara and Oana3} also demonstrated various factors (i.e. genetic, organic, and epilepsy-related) associated with the development of interictal psychosis in patients with epilepsy.

Psychoses in patients with any central nervous system (CNS) adversity, not only epilepsy but also other brain disorders, can be diagnosed as organic psychosis. The international criteria for mental disorders, either the ICD–10 or the DSM–IV, recognise the traditional dichotomy, i.e. functional and organic psychosis. However, since such CNS adversities are not invariably associated with the development of psychotic states, other additional conditions are required to generate psychotic symptoms. It is known that psychoses after brain injury occur more frequently in people with a family loading of psychoses. ^{Reference Corcoran and Malaspina4} Thus, individual (possibly constitutional) vulnerability to psychosis can be considered as a contributing factor to the development of organic psychosis and its severity.

As for classification systems for mental disorders, many limitations of the Kraepelinian dichotomy between schizophrenia and affective disorders have been discussed. ^{Reference Craddock and Owen5} Likewise, there appear to be limitations to the dichotomous view of organic and non-organic. The concept of organic psychosis has been useful to classify and treat patients, but it appears too simplistic to explain complex pathogenesis in such patients. It may be time to reconceptualise psychoses in patients with or without diagnosable brain disorders.