To test the utility of the UCLA posttraumatic stress disorder (PTSD) Index for DSM-IV (U-PTSD-I)(c) for predicting PTSD diagnosis in children with and without medical illnesses. The U-PTSD-I and a standard psychiatric interview were administered to medically ill children (n = 76) and children who experienced other traumatic events (n = 31). We found U-PTSD-I's sensitivity and specificity was better in the nonmedical illness cohort. Only intrusion symptoms were significantly associated with the diagnosis of PTSD in the medically ill. In conclusion, the U-PTSD-I performs better among general trauma versus medically ill patients. Intrusion symptoms should be focused on when assessing PTSD in medically ill children.