•BDNF and NGF plasma levels cannot serve as biomarkers of a second relapse in FES.•BDNF and NGF cannot be considered reliable biomarkers of state changes in schizophrenia.•There are significant correlations between BDNF and NGF and symptom severity.•BDNF and NGF serve as biomarkers of the underlying illness trait in later phases after a FES. Neurotrophins have been proposed to be involved in biological mechanisms which might underlie different clinical outcomes in schizophrenia. The aims of the present study were to examine the BDNF/NGF plasma levels in a cohort of first-episode schizophrenia (FES) patients in remission as potential biological predictors of relapse; to study the associations between these neurotrophins and the symptomatology severity through different stages after a FES in two independent cohorts. 2EPs-Cohort: 69 first-episode in clinical remission were included. BDNF/NGF plasma levels and symptom severity were measured at enrollment and at 3-year or at the time of the second episode/relapse. FLAMM-PEPs-Cohort: 65 first-episodes were also included. BDNF/NGF and symptom severity were obtained at enrollment and 2-year follow-up. Symptomatology was assessed with the Marder-PANSS-Factor scores. Plasma neurotrophins did not differ significantly over time and neither BDNF/NGF were predictors of relapse. Besides, in remission stages, baseline BDNF levels showed significant correlations with both positive and negative symptoms (p<0.05); NGF, with negative symptomatology (p<0.01). Similarly, in the FLAMM-PEPs-Cohort, baseline BDNF/NGF levels showed significant correlations with negative symptoms (and not positive symptomatology) at follow-up (p<0.05). In both cohorts, lower levels correlated with higher symptom severity. Findings did not support a role for BDNF/NGF plasma levels as biomarkers of relapse in FES patients. Nevertheless, baseline BDNF/NGF may lead to be considered potentially useful biomarkers of long-term severity in schizophrenia and of the underlying illness traits, specially of negative symptomatology severity. More longitudinal studies in FES samples and adding a control group are warranted to replicate these findings.