PURPOSE: This study was to evaluate the effect of arterial embolization on survival in patients with pelvic bone fractures and arterial bleeding. METHODS: From January 2001 to December 2007, in all, 18 patients with pelvic bone fractures that had been treated with interventional arterial embolization were included in this retrospective study. The Injury Severity Score (ISS), the Revised Trauma Score (RTS), the initial hemodynamic status, the blood gas analysis, blood transfusion data, and mortality were the main outcome measurements. RESULTS: Pelvic bone fractures were classified into lateral compression (LC), antero-posterior compression (APC), vertical shear (VS), and combined (CM) type according to the Young-Burgess classification. The Survivor group included 11 patients (61.1%), and the non-survivor group included 7 patients (38.9%). The mean ages for the survivor and the non-survivor groups were 40.0 and 45.6 years (p=0.517). The types of pelvic bone fractures were LC 11 (61.1%), APC 6 (33.3%), and VS 1 (5.6%): LC 7 (63.6%), and APC 4 (36.4%) in the survivor group and LC 4 (57.1%), APC 2 (28.6%), and VS 1 (14.3%) in the non-survivor group. The internal iliac artery was the predominant injured vessel among both the survivors (n = 5, 45.5%) and the non-survivors (n = 4, 57.1%). No differences in initial blood pressures, ISS, and RTS existed between the two groups, but the arterial pH was lower in the non-survivor group (pH 7.09 (+/-0.20) vs 7.30 (+/-0.08), p=0.018). The number of transfused 24-hour units of packed RBC was greater in the non-survivor group (24.1+/-12.5 vs 14.4+/-6.8, p=0.046). CONCLUSION: No differences in initial blood pressure and trauma scores existed between survivors and nonsurvivors with pelvic bone fractures, who had been treated with arterial embolization, but arterial pH was lower the in non-survivors.