Background. Deficits in a patient's ‘theory of mind’ (TOM) have been proposed to lead to psychosis. However, it remains unclear whether TOM deficits constitute a trait- or a state-related deficit and whether they respond to antipsychotic treatment, and also whether the change in TOM and change in psychosis are associated. Method. In the cross-sectional component of this study, 71 patients with psychotic disorders were included and TOM ability was measured using a hinting task in which subjects had to infer real intentions behind indirect speech. In the longitudinal study, a different cohort of 17 drug-free patients were included wherein they received antipsychotic treatment for 6 weeks and the effect on psychotic symptoms and TOM was measured every 2 weeks. Associations between TOM and psychopathology were assessed and a mixed effects model was used to investigate the rate of change over time. Results. Positive and Negative Syndrome Scale (PANSS) total scores were significantly associated with TOM scores. The hinting task was not associated with positive symptoms but was significantly associated with negative and general symptoms. The longitudinal arm of the study showed that both PANSS positive scores and TOM improved after medication was started, particularly during the first 2 weeks of antipsychotic treatment, but these changes were not associated. The TOM response at 2 weeks of antipsychotic treatment reached similar values to those obtained in the cross-sectional sample. Conclusions. Although TOM and psychotic symptoms are related to each other, antipsychotic treatment impacts each independently, suggesting a dissimilar cognitive or neurobiological substrate for the two.