BACKGROUND: A range of outcome measures has been developed to assess psychotic symptoms. Many of these are observer based assessments which rate the presence and severity of broad symptom types. There is a need for service-user generated multi-dimensional scales to assess psychosis. AIMS: To investigate the psychometric properties of a service user generated, self-report scale to assess recovery in relation to psychotic symptoms. METHOD: The reliability and validity of the Subjective Experiences of Psychosis Scale (SEPS) was investigated with 100 participants experiencing psychosis against observer rated assessments of psychosis and self report measures of affect, esteem, recovery and functioning. RESULTS: Two factors emerged representing positive and negative aspects of psychotic experiences. Test-retest reliability, internal consistency and sensitivity to change were good and the scales correlated with the PANSS, PSYRATS and measures of affect, esteem, recovery and functioning. CONCLUSIONS: The SEPS is a reliable and valid tool which can be used to evaluate outcome from treatment and which reflects the multi-dimensional experience of psychosis.