This study was undertaken in order to determine whether screening of viremic blood donations by testing of pooled donor samples could constitute a technically feasible transfusional safety measure. A pilot study of real-time simulation, on a day-to-day basis, of screening of three viral genomes (hepatitis B virus (HBV), hepatitis C virus (HCV) and human immunodeficiency virus (HIV)) was conducted by five French Blood Centers on plasma samples collected from blood donors and studied within undiluted samples and within sample pools of various sizes. This study was carried out within time conditions compatible with the release of platelets. For the detection of HCV and HIV genomes, the five laboratories achieved a sensitivity that decreased with the size of the sample pool. Four were successful in detecting all undiluted samples. In the 1/10 diluted samples, four failed to detect one HIV or HCV sample. In the 1/100 diluted samples, all laboratories failed to detect one or more HIV or HCV samples. For HBV genome, no participating laboratories detected all of the samples of the panel, even undiluted samples, and the sample pooling considerably affected sensitivity. The improvement and standardization of assays needs to be attained, and training of laboratories appears to be a step crucial for routine screening of viral genomes in blood donations.