Introduction. Bentall and colleagues (Bentall & Kaney, 1996; Kinderman & Bentall, 1996, 1997) claim that persecutory delusions are constructed defensively, for the maintenance of self-esteem. A central prediction of their model is that such delusions will be associated with discrepancies between overt and covert self-esteem. Methods. The present study employed a new methodology that has been widely used in investigations of implicit attitudes, the Implicit Association Test (IAT; Greenwald, McGhee, & Schwartz, 1998), to assess covert self-esteem and to test the above prediction. Overt self-esteem was assessed using the Rosenberg Self-Esteem Scale and an adjective self-relevance ratings measure. These measures were administered to 10 patients with acute persecutory delusions, 10 patients with remitted persecutory delusions, and 19 healthy control participants. Results. Patients with persecutory delusions were found to have lower covert self-esteem (as assessed using the IAT) than healthy controls and patients with remitted persecutory delusions. On two measures of overt self-esteem, however, the persecutory deluded group did not differ significantly from the other groups once the effects of comorbid depression had been taken into account. Conclusions. These results are thus consistent with a model of persecutory delusions as serving the defensive function of maintaining self-esteem.