Background: The Brain-derived Neurotrophic Factor (BDNF) modulates cognitive processes and is associated with increased risk of schizophrenia. Childhood trauma (CT) is frequent in patients with psychosis and severely affects course and outcome. Aims: We investigated the hypothesis that BDNF is associated with both CT and cognitive deficits in a sample of first-episode psychosis (FEP) cases and unaffected controls. Method: Participants with FEP and healthy controls were recruited between August 2008 and July 2011 from South London, UK. Childhood traumatic events were detected using the Childhood Experience of Care and Abuse Questionnaire (CECA-Q). Neuropsychological data were also collected. BDNF plasma levels were measured from fasting blood samples. Results: Data were available on 87 FEP patients and 152 controls. Our results showed a significant effect of separation (F = 5.5; df = 1,115; p = .02), physical (F = 4.7; df = 1, 118; p = .03) and sexual abuse (F = 5.4; df = 1,117; p = .02) on BDNF levels with lower levels among those who experienced the traumatic event compared to those who did not. Physical abuse predicted lower plasma levels of BDNF (beta = -.30; p = .03) whereas sexual and/or physical abuse showed a trend (beta = -.26; p = .06) in FEP patients but not in unaffected controls. No association between BDNF plasma levels and cognitive functions was found among patients with FEP and controls. Conclusion: Our findings suggest the possible involvement of BDNF in the onset of first-episode psychosis in individuals exposed to early trauma and propose BDNF as a potential clinical biomarker to detect the detrimental effects of CT on human brain plasticity. Crown Copyright