Design. A 1-week, prospective momentary assessment study. Methods. Data were collected using the experience sampling method, a structured self-assessment diary technique. The sample consisted of 158 individuals who ranged across the paranoia continuum. Participants with a psychotic disorder were recruited from in-patient and out-patient mental health services. Participants without psychotic disorder were sampled from the general population. Results. Specific aspects of emotional experience were implicated in the onset and persistence of paranoid episodes. Both an increase in anxiety and a decrease in self-esteem predicted the onset of paranoid episodes. Cross-sectionally, paranoid episodes were associated with high levels of all negative emotions and low level of self-esteem. Initial intensity of paranoia and depression was associated with longer, and anger/irritability with shorter duration of paranoid episodes. Conclusions. Paranoid delusionality is driven by negative emotions and reductions in self-esteem, rather than serving an immediate defensive function against these emotions and low self-esteem. Clinicians need to be aware of the central role of emotion-related processes and especially self-esteem in paranoid thinking.