Existing self-rating scales to measure depersonalization either show dubious face validity or fail to address the phenomenological complexity of depersonalization. Based on a comprehensive study of the phenomenology of this condition, a new self-rating depersonalization questionnaire was constructed. The Cambridge Depersonalization Scale is meant to capture the frequency and duration of depersonalization symptoms over the 'last 6 months'. It has been tested on a sample of 35 patients with DSM-IV depersonalization disorder, 22 with anxiety disorders, and 20 with temporal lobe epilepsy. Scores were compared against clinical diagnoses (gold standard) and correlated with the depersonalization subscale of the Dissociation Experiences Scale (DES). The scale was able to differentiate patients with DSM-IV depersonalization disorder from the other groups, and showed specific correlations with the depersonalization subscale of the DES (r=0.80; P=0.0007). The scale also showed high internal consistency and good reliability (Cronbach alpha and split-half reliability were 0.89 and 0.92, respectively). The instrument can, therefore, be considered as valid and reliable, and can be profitably used in both clinical and neurobiological research.