Several CT scan scores have been proposed to discriminate between splenic injuries requiring surgical intervention, and those that may be treated conservatively. The aim of this study is to compare three selected CT scan scores (Buntain, Mirvis, Resciniti) in their ability to correctly define the proper course of therapy following splenic injury. 44 adult patients presenting an isolated splenic trauma, confirmed by CT scan on admission and treated solely according to clinical criteria, have been studied. The 44 CT scans were retrospectively classified according to the three CT scan scoring systems. Correlation with the treatment performed allows the specificity and sensitivity of each score to be calculated and compared. The best score (Resciniti) uses a combination of parenchymal and intra-abdominal liquid scores. The use of such a scoring system alone to dictate therapy would have resulted in either an increased rate of surgical intervention, or an unacceptable rate of failure of conservative treatment. However, the determination of the expected rate of failure of conservative treatment can be useful in determining the level and frequency of serial surveillance examinations.