Mental health services and research have been dominated for several decades by a rather simplistic, reductionistic focus on biological phenomena, with minimal consideration of the social context within which genes and brains inevitably operate. This 'medical model' ideology, enthusiastically supported by the pharmaceutical industry, has been particularly powerful in the field of psychosis, where it has led to unjustified and damaging pessimism about recovery. The failure to find robust evidence of a genetic predisposition for psychosis in general, or 'schizophrenia' in particular, can be understood in terms of recently developed knowledge about how epigenetic processes turn gene transcription on and off through mechanisms that are highly influenced by the individual's socio-environmental experiences. To understand the emerging evidence of the relationship between adverse childhood events and subsequent psychosis, it is necessary to integrate these epigenetic processes, especially those involving the stress regulating functions of the HPA axis, with research about the psychological mechanisms by which specific types of childhood trauma can lead to specific types of psychotic experiences. The implications, for research, mental health services and primary prevention, are profound.