Even after the implementation of the nucleic acid amplification testing (NAT) system, there remains a residual risk of viral transmission through blood transfusion because of the limited sensitivity of the reagents used and the pooling strategy of the current NAT system. From the calculation using NAT yield and the length of the window period, we presume that we will obtain 0.75 donations for human immunodeficiency virus and 0.58 donations for hepatitis C virus annually that are individual donation-NAT positive but 50-individual pool-NAT negative, figures that are comparable with those in other developed countries. The number of donations potentially positive for the hepatitis B virus genome is, however, considerably high in Japan and is estimated to be more than 100 annually, which is the sum of the donors in the minipool-NAT window period and the chronic carriers with a low viral load. The incidence of bacterial sepsis after transfusion is relatively low in Japan. This incidence is possibly attributable to the short shelf lives of platelet concentrate and red blood cell component, which are 3 and 21 days, respectively. In Japan, the implementation of a new technology to screen out or abrogate infectious agents in blood components is necessary while considering the balance between benefits and possible new risks or costs.