The current classification of the psychoses remains based on Kraepelin’s division into two distinct diagnostic categories: dementia praecox and manic-depressive insanity with the older terms being replaced by schizophrenia (SZ) and bipolar disorder (BD). The limitations of this dichotomous approach including overlapping clinical features, diagnostic instability and lack of a symptomatic point of rarity have been apparent for many years. In this chapter, I review recent genetic and genomic data as well as findings from studies of cognition and neuroimaging that further challenge the validity of the dichotomous classification. I also address the question of what biological insights have emerged from genomic and other studies. This review suggests the following conclusions. The current dichotomous classification does not define conditions that are distinct or homogeneous whether regarding psychopathology, aetiology or mechanism. Rather, the evidence suggests a continuum with several dimensions. One dimension is a gradient of neurodevelopmental impairment which is greatest in cases currently diagnosed with SZ, followed by schizoaffective disorder, then by BD. This is indexed by cognitive impairment and particularly the degree to which cognitive function deviates from that expected based on family history. In addition, current evidence suggests that the clinical picture expressed by an individual with a psychotic disorder is the result of a confluence of partly orthogonal, transdiagnostic symptom dimensions and their underlying genetic risk factors. Finally, psychotic disorders are fundamentally disorders of neuronal function and development across multiple brain regions with broad impacts on brain function. The pattern of symptoms, cognitive impairments and other features in each case likely reflects the pattern of resulting dysfunction in specific brain regions or circuits depending on the individual mix of genetic and other risk or modifying factors.