Within the embryonic early psychosis field in the early 1990s, the conceptualisation and definition of an at-risk or ultra-high-risk (UHR) mental state for psychosis was a breakthrough which transformed the clinical and research landscape in psychiatry. Twenty-five years later, we have a new evidence base that has illuminated the neurobiology of the onset phase of psychotic disorder, delivered Cochrane level 1 evidence showing that the onset of full-threshold sustained psychotic disorder can be at least delayed, and is paving the way to a new generation of transdiagnostic research. Here, we document the contribution of the UHR approach to understanding the underlying mechanisms of psychosis onset as well as the long-term outcomes. Particularly, we highlight that psychosis onset can be delayed in those meeting UHR criteria and that these criteria have a higher valence for subsequent psychotic disorders and some valence for persistent non-psychotic syndromes. Critiques have helped to identify some of the limitations of this paradigm, which are acknowledged. These include evidence that psychotic disorders can emerge more acutely and from other, as yet undefined, precursor states. Rather than defending, or alternatively questioning the value of, the UHR approach, we propose a broader, transdiagnostic staging model that is consistent with the pluripotent and variably comorbid trajectories for mental disorders. This approach moves beyond psychosis to capture a wider range of subthreshold symptoms and full-threshold disorders, thus enhancing prediction for the emergence and progression of a range of mental disorders, as well as providing new avenues for early intervention and prevention.