Thromboelastography or rotational thromboelastometry, is being increasingly utilized in cardiac surgery of late. However, it is an indirect test and is not available in all centers. Low fibrinogen levels before and after cardiopulmonary bypass (CPB) have been described to be associated with postoperative bleeding in cardiac surgery. This study explored the usefulness of reduction ratio of the fibrinogen levels before CPB (preCPB) and after CPB (postCPB) in predicting postoperative hemorrhage. A retrospective, observational study of adult patients who underwent cardiac surgery with CPB between February 2014 and January 2016 was conducted, which included a total of 264 patients. The fibrinogen levels were measured twice, preCPB and postCPB, and the fibrinogen reduction ratio was acquired [(preCPB − postCPB)/preCPB]. Postoperative blood loss, which was defined as the blood collected from the chest drain for 12 hours following arrival at the intensive care unit, was considered severe if it was more than 1,000 mL. A multivariate analysis showed that fibrinogen reduction ratio, sex, and postCPB platelet count were significantly associated with severe postoperative bleeding. However, the pre- and postCPB fibrinogen levels were not significantly associated with severe bleeding. Furthermore, a fibrinogen reduction ratio of > 41.3% was independently associated with postoperative severe bleeding, with an odds ratio of 3.472 (1.483–8.162). These results suggest that the reduction ratio of pre- and postCPB fibrinogen levels may be utilized in predicting postoperative bleeding.