The neurodevelopmental hypothesis of schizophrenia represents the disorder as an expression of an alteration during the brain development process early in life. Neurodevelopmental variables could become a trait marker, and the study of these variables in children and adolescents at clinical high risk for psychosis (CHR) could identify a specific cluster of patients who later developed psychosis. The aim of this study is to describe clinical and neurodevelopment predictors of transition to psychosis in child and adolescent participants at CHR.Naturalistic longitudinal two-center study of 101 CHR and 110 healthy controls (HC) aged 10-17. CHR participants were children and adolescents aged 10-17, meeting one or more of the CHR criteria assessed at baseline and at 18 months' follow-up. Neurodevelopmental variables assessed were obstetric complications, delay in principal development milestones, and presence of a neurodevelopment diagnosis. Pairwise comparisons, linear regressions, and binary logistic regression were performed.A transition rate of 23.3% at 1.5 years was observed. Participants who developed psychosis (CHR-P) showed higher rates of grandiosity and higher proportions of antipsychotic medication intake at baseline compared to participants who did not develop a psychotic disorder (CHR-NP). In terms of neurodevelopment alterations, CHR-P group showed a higher proportion of participants reporting delay in language development than the CHR-NP and HC groups. The odds of psychosis increased by 6.238 CI 95% [1.276-30.492] for a one-unit increase in having a positive score in grandiosity; they increased by 4.257 95% CI [1.293-14.023] for a one-unit increase in taking antipsychotic medication, and by 4.522 95% [1.185-64.180] for showing language development delay. However, the p-values did not reach significance after adjusting for multiple comparisons.A combination of clinical and neurodevelopmental alterations could help predict the transition to psychotic disorder in a CHR child and adolescent sample. Our results suggest the potential utility of collecting information about neurodevelopment and using these variable multifactorial models to predict psychosis disorders.