Background <p>Aggression is a major clinical issue in schizophrenia, but its biological basis remains unclear. This study examined the relationship between thyroid hormones, psychiatric symptoms, and aggression in drug-naïve schizophrenia (DNS) patients using causal inference methods.</p> Methods <p>In a case–control study, we examined 161 DNS patients and 63 healthy controls. Aggression was assessed with the Modified Overt Aggression Scale (MOAS) and psychiatric symptoms with the Positive and Negative Syndrome Scale (PANSS). Serum thyroid-stimulating hormone (TSH), triiodothyronine (T3), and thyroxine (T4) were measured. DAG modeling, multivariable regression, mediation, and sensitivity analyses were conducted to characterize the relationships among thyroid hormones, psychiatric symptoms, and aggression.</p> Results <p>DNS patients exhibited significantly lower TSH levels than controls (<i>t</i> = − 24.81, <i>p</i> &lt; 0.001). Within the schizophrenia cohort, the high-aggression group showed higher T3 levels (<i>t</i> = 2.24, <i>p</i> = 0.03) and more severe positive symptoms (<i>t</i> = 2.85, <i>p</i> &lt; 0.001) than the low-aggression group. DAG analysis identified a direct association between T3 and aggression severity. In multivariable regression analyses, T3 remained significantly associated with aggression severity after adjustment for T4, age, and sex (<i>β</i> = 5.08, <i>p</i> = 0.01); and the association remained significant after further adjustment for positive and negative symptom dimensions (<i>β</i> = 4.99, <i>p</i> = 0.01). Mediation analyses found no evidence that positive or negative symptom dimensions explained this association. Sensitivity analyses supported the robustness of the findings.</p> Conclusions <p>Higher T3 levels were associated with greater aggression severity in patients with schizophrenia.</p>