Background: Adequate care of trauma patients following traffic accidents requires early and needs-based deployment of appropriate rescue resources, as well as shock room treatment that is appropriate to the expected pattern of injuries. Therefore, validated discriminants with the highest possible predictive accuracy are necessary for alerting all members of the rescue chain. The current version of the S3 guideline for polytrauma/severely injured patient treatment lists trauma-related shock room activation criteria, which are re-evaluated in this study using the German In-Depth Accident Study (GIDAS) database for traffic accidents. Methods: In the present descriptive, retrospective registry study, 42,310 traffic accidents involving 57,837 car occupants were analyzed using the GIDAS database (01/1999–06/2021), and parameters for predicting severe injuries (AIS 3+) were collected. Results: The parameters “Intrusion of more than 50–70 cm” (PPW 48–59%), “Speed change > 30 km/h” (PPW 10%), and “Pedestrian/cyclist collision” (PPW 8%) showed low predictive accuracy or are not preclinically assessable. In contrast, “Fatality of an occupant” (PPW 45%) and “Ejection from the vehicle” (PPW 57% for ejected individual, PPW 28% for co-occupants) demonstrated good predictive accuracy. Furthermore, the parameter “Vehicle deformation” exceeding 130 cm (frontal impact, PPW 33%) and 100 cm (side impact, PPW 24%) exhibited good predictive accuracy and could contribute to the improvement of adequate shock room activation. Conclusion: In our study, we identified discriminants with both good and poor predictive accuracy for severe injuries. These should form the basis for the activation of a shock room team B (partial alarm) and be included in the upcoming version of the S3 guideline for polytrauma/severely injured patient treatment. In addition, further parameters should be examined and provided to emergency dispatchers through query models.