The presented study was initiated to develop a scoring system for the prediction of red blood cell transfusion requirement in the early care of trauma patients. All trauma patients admitted to our institution who needed trauma team activation were evaluated during a 4-year period. A set of nine parameters with possible predictive value for the need of blood transfusion was recorded. All relevant data can be acquired during the first 10 min in the emergency room (ER). The data underwent multivariate logistic regression analysis for correlation and the calculation of predictive power. To transform the model into a practical score, we rounded all coefficients. The predictive power of the score was evaluated based on a linear regression equation. Of the 1103 patients (Injury Severity Score [ISS] 21 +/- 16) included in the study, 116 (10.5%; ISS 39 +/- 18) received blood in the ER. Early transfusion need was significantly correlated with systolic blood pressure (SBP) <90 mmHg (coefficient 2.5), SBP 90-120 mmHg (1.5), free fluid in abdominal ultrasound (2.0), clinically unstable pelvic ring fracture (1.5), age 20-60 years (0.5), age >60 years (1.5), admission from scene (1.0), traffic accident (1.0) and fall from >3 m (1.0). The probability for transfusion exponentially increased with the sum of points in the ER transfusion score, i.e. from 0.7% at one point to 5% at three points and 97% at 9.5 points maximum. To establish a practical cutoff point (risk <5%) a low-risk group was defined at <points (64% of the whole study group). The presented ER transfusion score is based on rapidly assessable parameters. The score identifies patients in need for immediate red blood cell substitution. Cost effectiveness appears to be a further advantage of the score. For patients not in need of urgent transfusion (low-risk group), the costs for transportation, cross-matching and loss by maltreatment of blood products may be avoided.