This study evaluates whether baseline deviations from a normative reference range in subcortical volumes are better predictors of antipsychotic treatment response than raw volumes in patients with first-episode psychosis who were naive to antipsychotic medication.