OBJECTIVE The notion of a disordered self as a core disturbance of schizophrenia was proposed in many foundational texts. Recent studies, spurred by the development of the Examination of Anomalous Self-Experience (EASE), seem to indicate that self-disorders are a specific manifestation of schizophrenia vulnerability. Follow-up studies of help-seeking, prodromal and first-admission patients have demonstrated the utility of self-disorders for predicting later schizophrenia-spectrum disturbance. We wished to extend these findings by gauging the predictive value of self-disorders in a premorbid, non-clinical population at high risk for schizophrenia. METHODS Children from the Copenhagen High-Risk Project with high-genetic risk for schizophrenia (N = 212) were assessed premorbidly (average age = 15), and diagnostically re-evaluated after 10 and 25 years. Since the EASE was not available at the time of premorbid assessment, we hypothesized that a proxy scale drawn from the Minnesota Multiphasic Personality Inventory (MMPI) could distinguish those who later developed a schizophrenia-spectrum disorder (N = 68) from those who remained healthy (N = 64). The Self-Disorder Scale comprised 32 items whose content suggested an aspect of self-disorder as measured by the EASE. RESULTS Premorbid Self-Disorder Scale scores significantly predicted lifetime schizophrenia-spectrum diagnosis in the high-risk cohort. Although there was considerable item overlap between the new scale and an existing MMPI scale (psychoticism), the overlap did not account for the Self-Disorder Scale's predictive efficacy. CONCLUSIONS The results support the notion of self-disorders as a core vulnerability feature in schizophrenia, detectable premorbidly in those developing later schizophrenia-spectrum disorders.