To study the psychometric properties of the Childhood Trauma Questionnaire-Short Form (CTQ-SF), we determined its dimensional structure, measurement invariance across presence of emotional disorders, the association of the CTQ-SF with an analogous interview-based measure (CTI) across presence of emotional disorders, and the incremental value of combining both instruments in determining associations with severity of psychopathology. The sample included 2,308 adults, ages 18-65, consisting of unaffected controls and chronically affected and intermittently affected persons with an emotional disorder at Time 0 (T0) or 4 years later at T4. Childhood maltreatment was measured at T0 with an interview and at T4 with the CTQ-SF. At each wave, patients were assessed for Diagnostic and Statistical Manual of Mental Disorders (4th ed., or DSM-IV; American Psychiatric Association, 1994)-based emotional disorders (Composite Interview Diagnostic Instrument) and symptom severity (Inventory of Depressive Symptomatology, Beck Anxiety Inventory, Fear Questionnaire). Besides the correlated original 5-factor solution, an indirect higher order and direct bifactorial model also showed a good fit to the data. The 5-factor solution proved to be invariant across disordered-control comparison groups. The CTQ-SF was moderately associated with the CTI, and this association was not attenuated by disorder status. The CTQ-SF was more sensitive in detecting emotional abuse and emotional neglect than the CTI. Combined CTQ-SF/CTI factor scores showed a higher association with severity of psychopathology. We conclude that although the original 5-factor model fits the data well, results of the hierarchical analyses suggest that the total CTQ scale adequately captures a broad dimension of childhood maltreatment. A 2-step measurement approach in the assessment of childhood trauma is recommended in which screening by a self-report questionnaire is followed by a (semi-)structured diagnostic interview.