Background: Childhood and adolescent trauma is a risk factor for developing psychosis-spectrum disorders. The current study aimed to assess how childhood trauma might predict psychosis symptomatology, and how patients' beliefs of whether trauma is the cause of psychosis might affect this association. Methods: Ninety-six first-episode psychosis patients were assessed for childhood traumatic experiences with the Brief Betrayal Trauma Survey, and for psychosis symptoms with the Positive and Negative Syndrome Scale. Results: Non-interpersonal trauma predicted higher positive symptoms, whereas more trauma domains experienced predicted lower negative symptoms. Almost half of the participants believed trauma to be related to psychosis, were 12 times more likely to reexperience trauma through psychosis, and had higher excitative and emotional symptoms. Non-interpersonal trauma also predicted higher positive symptoms in this group. Those who did not believe trauma to be the cause of psychosis had higher negative symptoms, and a negative dose-response was found for negative and disorganised symptoms, in which more trauma domains experienced predicted lower scores. Conclusions: Results imply two traumagenic pathways to psychosis, one characterised by positive, excitative, and emotional symptoms, and one negative subtype, characterised by negative and disorganised symptoms. Clinical implications for how findings might contribute to better treatments are discussed.