For the purpose of retrospective assessment and characterization of childhood trauma in adults, the factorial validity of the short form of the Childhood Trauma Questionnaire (CTQ-SF) in Germany was evaluated by conducting confirmatory factor analyses for three samples including 1,524 adult psychiatric patients, 224 inmates, and 295 university students. In addition, sex-specific confirmatory factor analyses were performed within each sample. Because several authors have suggested a different factor structure than that originally proposed in the manual, two competing models focusing on the Physical neglect subscale were examined. In psychiatric patients and inmates, the fit indices were reasonable to good. Among the students, factor loadings were markedly lower, and fit indices were poor. Sex-specific analyses did not indicate sex differences. Comparing the original and the alternative models revealed better fit indices of the original factor structure. The present findings indicate that the German version of the CTQ-SF has factorial validity in psychiatric patients and inmates, but not in students.