An emerging consensus declares the need to treat the onset of a psychotic illness as early as possible, but, numerous studies have found an extremely high number of psychotic experiences in the general population, which significantly exceeds the number of established schizophrenia. Detecting the initial symptoms is therefore primordial as it enables rapid treatment on which the long term prognosis partly depends. If in daily clinical medicine the psychotic episode is obvious to the physician, there are semiological presentations where the diagnosis is more arduous. This brings to light the difficulty of knowing if the psychotic symptom is not being confused with another symptom within a nosographic framework that is distinct from a psychosis. The characterisation of the psychotic symptom is at the heart of a problem of differentiating a non pathological psychotic experience from a psychotic symptom which marks a proven pathology. There is a continuity in the distribution of the psychotic experience in the general population. Just as an individual's non pathological experience can with time become a genuine symptom, a pathological element which took a while to manifest itself as such. Certain psychotic experiences will disappear or remain non pathological, whilst others will develop progressively into a schizophrenic pathology. The current psychometric tools do not appear to be sensitive enough or specific enough to establish a progressive distinction. However, new research is looking into the possibility of monitoring the evolution of psychotic experiences using functional cerebral imaging.