In order to test the trauma model of dissociation, the authors compared two samples with similar rates of reported childhood physical and sexual abuse: 502 members of the general population in Winnipeg, Canada, and 304 psychiatric outpatients at Shanghai Mental Health Center in Shanghai, China. There is virtually no popular or professional knowledge of dissociative identity disorder in China, and therefore professional and popular contamination cannot be operating. According to the trauma model, samples from different cultures with similar levels of trauma should report similar levels of dissociation. According to the sociocognitive model, in contrast, pathological dissociation is not related to trauma and should be absent in samples free of cultural and professional contamination. Of the 304 Chinese respondents, 14.5% reported childhood physical and/or sexual abuse compared to 12.5% of the Canadian sample. Both samples reported similar levels of dissociation on the Dissociative Experiences Scale and the Dissociative Disorders Interview Schedule. The findings support a specific prediction of the trauma model of dissociation not tested in previous research, and are not consistent with the sociocognitive, contamination or iatrogenic models of dissociative identity disorder.