Background <p>Biological predictors of relapse in psychosis remain underexplored. We examined whether dopamine synthesis capacity (DSC) predicts psychotic relapse one year after remission, and its associations with baseline clinical and cognitive factors.</p> Study Resign <p>In a prospective one-year follow-up study, forty-seven patients with remitted psychosis and fifteen age- and sex-matched healthy controls underwent positron emission tomography/computed tomography (PET/CT) scanning at baseline, with DSC indexed by mean standardized uptake value ratios (SUVr). Relapse was assessed monthly over one year.</p> Study Results <p>SUVr measures did not differ between patients and controls. Within patients, a higher putamen-to-cerebellum SUVr ratio at baseline was associated with male sex (<i>p</i> = 0.019), unmarried status (<i>p</i> = 0.044), more premorbid schizoid-schizotypal traits (<i>p</i> = 0.015), and greater general psychopathology (<i>p</i> = 0.010); only the latter remained significant after correction for multiple comparisons. Higher caudate-to-cerebellum SUVr was also linked to schizoid-schizotypal traits (<i>p</i> = 0.007), while higher putamen-to-caudate SUVr correlated with better insight (<i>p</i> = 0.017). At one year, relapsers showed lower putamen-to-caudate SUVr (<i>p</i> &lt; 0.001), were more often immigrants (<i>p</i> = 0.002), had poorer insight into antipsychotic benefits (<i>p</i> = 0.003), and were more likely to have discontinued medication at baseline (<i>p</i> = 0.029). Penalized logistic regressions confirmed that lower baseline putamen-to-caudate SUVr predicted relapse independently of medication status at both baseline (<i>p</i> = 0.037) and follow-up (<i>p</i> = 0.044).</p> Conclusions <p>Baseline reductions in striatal DSC predicted relapse regardless of medication, supporting its role as a marker of dopaminergic vulnerability.</p>