Purpose: To test predicted differences in phenomenology, irrational beliefs/demands and underlying insecurity of the two subgroups (poor me and bad me) of paranoia proposed by Trower and Chadwick (1995). Method: 40 inpatients suffering from persecutory delusions were classified as either poor me or bad me paranoid. P~cipants were assessed for depression (HADS: Hospital Anxiety and Depression Scale), self-esteem (RSE: Rosenberg Self-Esteem Inventory), irrational beliefs/demands (SGABS: Shortened Attitudes and Beliefs Scale), insecurity of the self (SOS: Selfand Other Scale) and negative person evaluations (EBS: Evaluative Beliefs Scale). Participants were also tested on a Dot Probe Task, which measured reaction time to a dot when it replaced poorme words, badme words and neutral words. Results: A number of the predictions were supported. Namely, poor me and badme groups displayed the predicted phenomenological differences (self-esteem, depression and negative person evaluations). The Dot Probe Task highlighted the predicted underlying insecurity of the poorme group but not the bad me group. The remaining questionnaires' measures of insecurity of the self and irrational beliefs/demands (SOS, SGABS), although non-significant, display some in:teresting findings. Conclusion: The data supports the predictions of the presence of two distinct topographies of paranoia; however, the associated underlying insecurities and associated underlying insecurities and irrational beliefs/demands are more complex. Further research is needed to explore these specific predictions.