While most standard symptom scales regard the 'psychotic' or 'positive' dimension of schizophrenia as a single factor, several lines of evidence suggest that psychosis itself is a multidimensional phenomenon. The foregoing literature suggested at least five distinct dimensions to psychosis; to test this, we developed, validated and applied an instrument to measure these dimensions and then applied it to examine the effect of antipsychotics on the different dimensions of the psychotic experience. The Dimensions of Psychosis Instrument (DIPI) was administered to 91 psychotic patients with schizophrenia spectrum disorders and a confirmatory factor analyses (CFA) was carried out to examine the five dimensions: cognitive preoccupation (CP) with the psychotic experience; emotional involvement (EM); behavioural impact (BI) of the experience; conviction (CO) in it; emotional; and external perspective (EP) about the experience. In a separate cohort of 17 prospectively treated patients, the impact of antipsychotics on these dimensions was assessed. BI showed the greatest improvement (32%) at 2 weeks, while CP and emotional improved somewhat less (22% and 14%, respectively). Improvement in CO was limited (6%) while EP showed no change. These results suggest that over the first few weeks of treatment, antipsychotics rapidly reduce the behavioural impact of the principal psychotic symptom and decrease cognitive and emotional preoccupation with it, without greatly altering the patients' conviction in or perspective about their psychotic experience.