Background Early recovery from trauma exposure in youth is poorly understood. This prospective longitudinal study examined the early course of traumatic stress responses in recently trauma‐exposed youth, evaluated the revised DSM‐5 acute stress disorder (ASD) and PTSD diagnoses and alternative diagnoses, and identified risk factors for persistent traumatic stress. Method Participants were 8‐ to 17‐year‐old emergency departments attendees exposed to single incident traumas. Structured clinical interviews were undertaken at 2 (n = 226) and 9 weeks (n = 208) posttrauma. Results Using the revised criteria in DSM‐5, 14.2% met criteria for ASD at week 2 and 9.6% met criteria for PTSD at week 9. These prevalences were similar to the corresponding DSM‐IV diagnoses (18.6% ASD at week 2; 8.7% PTSD at week 9). Using the same diagnostic criteria (DSM‐IV or DSM‐5) across assessments (i.e., “2‐week PTSD”) suggested that caseness declined in prevalence by approximately half. Overlap between DSM‐IV and DSM‐5 ASD and DSM‐5 preschool child PTSD diagnoses was considerable. Two diagnoses were strongly predictive of corresponding week 9 diagnoses. Youth with ASD who subsequently had PTSD reported more negative alterations in cognition and mood at 2 weeks than those youth who did not develop PTSD. Conclusions Youth exposed to single‐event traumas experience considerable natural recovery in the first months posttrauma. Using DSM‐5 criteria, ASD may not capture all clinically significant traumatic stress in the acute phase and is only moderately sensitive for later PTSD. Future research needs to address the role and etiology of negative alterations in cognition and mood symptoms.