INTRODUCTION: Non-operative management of blunt splenic injury in adults has been applied with increasing frequency. However, predictive criteria for successful non-operative management are still a matter of debate. METHODS: we retrospectively reviewed all cases of blunt splenic injury in adult patients from 1997 to 2006. RESULTS: Of 190 patients with blunt splenic trauma (median age: 33 years, range 16-98), 43.7% (n=83) underwent emergency surgical intervention (Group I), and 56.3% (n=105) of patients were admitted for conservative treatment of splenic trauma. Conservative treatment was successful in 76.6% (n=82) (Group II), while 23.4% (n=25) of patients required a laparotomy (Group III). Ultimately, 43.2% of patients were successfully managed non-operatively, and 56.9% underwent laparotomy. Mechanism of injury was not significantly different among three groups. Group I patients presented significantly more frequently with hypovolemic shock (p<0.01), associated injuries (p<0.01), and high grade of splenic injury (p<0.01). All patients with active bleeding as evidenced by extravasation on CT scan, underwent exploratory laparotomy. Failure of non-operative management increased significantly with splenic trauma grade (grade I (0%), grade II (22.6%), grade III (27.6%) and grade IV (40%), (p<0.01) and with quantity of hemoperitoneum (10.4% of patients with small, 22.2% of patient with moderate, and 47.8% with large hemoperitoneum). The median interval for conservative treatment failure was 3 days (range: 1-15). Splenic injuries were operatively controlled by splenectomy (91.6%) and splenorrhaphy (8.4%). CONCLUSION: Suitability of adult patients with blunt splenic injury for non-operative management may be predicted at initial presentation, based on hemodynamic status and associated injuries. The quantity of hemoperitoneum and magnitude of splenic injury are predictive factors for failure of conservative treatment. Early definition of these factors may help identify those patients likely to be successfully treated without laparotomy.