The main goal of the present study was to explore the latent structure of schizotypy as an indicator of psychosis liability, in a community-derived sample of adolescents. Links to mental health difficulties, prosocial behaviour, suicidal ideation, bipolar-like experiences and psychotic-like experiences (PLEs) (severity and distress) were compared across schizotypy latent profiles. The present research included 1588 adolescents selected by a stratified random cluster sampling. The Oviedo Schizotypy Assessment Questionnaire (ESQUIZO-Q), The Paykel Suicide Scale (PSS), The Strengths and Difficulties Questionnaire (SDQ), The Prodromal Questionnaire-Brief (PQ-B), The Mood Disorder Questionnaire (MDQ), The Penn Matrix Reasoning Test (PMRT), The Family Affluence Scale-II (FAS-II), and The Oviedo Infrequency Scale (INF-OV) were used. Using latent profile analysis four latent classes (LC) were identified: "Positive schizotypy" (14.1%, n = 224), "Low schizotypy" (51.9%, n = 825), "Social Disorganization schizotypy" (27.2%, n = 432), and "High schizotypy" (6.7%, n = 107). The "High schizotypy" class scored higher on several psychometric indicators of psychopathology (ie, mental health difficulties, suicide ideation, bipolar-like experiences and PLEs) relative to the other three LC. Four groups of adolescents with different patterns of schizotypal traits and different clinical-pathological meaning were found. Deficits found across schizotypy latent profiles, resembling those found in patients with psychosis and ultra-high risk samples. The identification of homogeneous subgroups of adolescents potentially at risk for psychosis may help us in the prevention of psychotic-spectrum disorders and mental health problems.