PURPOSE: To determine prognostic factors and evaluate outcomes of transcatheter arterial embolization in severely injured patients in hemodynamically unstable condition with multicompartmental bleeding.¦MATERIALS AND METHODS: Between June 2000 and May 2008, 36 consecutive patients treated with transcatheter arterial embolization for major retroperitoneal bleeding associated with at least one additional source of bleeding were retrospectively reviewed. Mean Injury Severity Score (ISS) was 49.4 ± 15.8. Univariate and multivariate analyses were performed to identify parameters associated with failure of embolization, need for additional surgery to control bleeding, and fatal outcome at 30 d.¦RESULTS: Embolization was technically successful in 35 of 36 patients (97.2%) and resulted in immediate and sustained (> 24 h) hemodynamic improvement in 29 (80.5%). Additional hemostatic surgery was necessary after embolization in six patients (16.6%). Fifteen patients (41.6%) died within 30 d. Failure to restore hemodynamic stability was correlated with the rate of administration of packed red blood cells (P = .014), rate of administration of fresh frozen plasma (FFP; P = .031), and systolic blood pressure (SBP) immediately before embolization (P = .002). The need for additional surgery was correlated with FFP administration rate before embolization (P = .0002) and hemodynamic success (P = .003). Death was correlated with Glasgow Coma Scale score at admission (P = .001), ISS (P = .014), New Injury Severity Score (P = .016), number of injured sites (P = .012), SBP before embolization (P = .042), need for vasopressive drugs before embolization (P = .037), and hemodynamic success (P = .0004).¦CONCLUSIONS: In patients in hemodynamically unstable condition, transcatheter arterial embolization effectively controls bleeding and improves hemodynamic stability. Immediate survival is related to hemodynamic condition before embolization, and 30-d mortality is mainly related to associated brain trauma.