Background: During the initial assessment of patients with potential severe injuries, radiological examinations are performed in order to rapidly diagnose clinically relevant injuries. Previous studies have shown that performing these examinations routinely is not always necessary and that trauma patients are exposed to substantial radiation doses. The aim of this study was to assess the amount and findings of radiological examinations during the initial assessment of trauma patients and to determine the radiation doses to which these patients are exposed to. Methods: We analyzed the 1124 patients included in a randomized trial. All radiological examinations during the initial assessment (i.e., primary and secondary survey) were assessed. The examination results were categorized as positive findings (i.e., (suspicion for) traumatic injury) and normal findings. The effective radiation doses for the examinations were calculated separately for each patient. Results: Eight hundred and three patients were male (71 %), median age was 38 years, and 1079 patients sustained blunt trauma (96 %). During initial assessment, almost 3900 X-rays were performed, of which 25.4 % showed positive findings. FAST of the abdomen was performed in 989 patients (88 %), with positive findings in 10.6 %. Additional CT scanning of specific body regions was performed 1890 times in 813 patients (72.1 %), of which approximately 43.4 % revealed positive findings. Hemodynamically stable patients showed more normal findings on the radiographic studies than unstable patients. The mean radiation doses for the total population was 8.46 mSv (±7.7) and for polytraumatized patients (ISS ≥ 16) 14.3 mSv (±9.5). Conclusion: Radiological diagnostics during initial assessment of trauma patients show a high rate of normal findings in our trauma system. The radiation doses to which trauma patients are exposed are considerable. Considering that the majority of the injured patients are hemodynamically stable, we suggest more selective use of X-ray and CT scanning.