As research into psychotic illness evolves along established lines, insights are emerging that deviate from those lines and challenge more fundamentally our understanding. On the background of a new generation of studies on first-episode psychosis, investigations across the gene–environment interface and the intersection with ‘normal’ human mentation heighten these concerns. Using findings from the Cavan-Monaghan First Episode Psychosis Study (CAMFEPS) as an exemplar, we here review the complexity of these challenges from the perspective of this real-world setting. They range from trans-diagnostic epidemiology and clinical characterisation, through molecular genetics, social milieu, developmental pathobiology and functional outcome across arbitrary diagnostic boundaries, to the evidence base for early intervention and more radical conceptualisations and structures for provision of mental health care.