The authors examined the Trauma Symptom Inventory's (TSI) ability to discriminate 88 student post-traumatic stress disorder (PTSD) simulators screened for genuine PTSD from 48 clinical PTSD-diagnosed outpatients. Results demonstrated between-group differences on several TSI clinical scales and the Atypical Response (ATR) validity scale. Discriminant function analysis using ATR revealed 75% correct patient classification but only 48% correct simulator classification, with an overall correct classification rate of 59% (positive predictive power [PPP] = .71; negative predictive power [NPP] = .51). Individual ATR cutoff scores did not yield impressive classification results, with the optimal cutoff (T score = 61) correctly classifying only 61% of simulators and patients (PPP = .66, NPP = .54). Although ATR was not developed as a malingered PTSD screen, instead serving as a general validity screen, caution is recommended in its current clinical use for detecting malingered PTSD.