In 1994, quarantine fresh-frozen plasma (Q-FFP) was introduced in Germany in order to reduce the risk of HIV and HCV transmission. In 1998, an acute HCV infection of a patient was reported to us. The look-back revealed that this patient had received two Q-FFP from a donor who had seroconverted for HCV in the meantime. Recipients of further plasma donations from this donor were identified. Back-up specimens of these donations were investigated in several laboratories. A total of 25 additional HCV-PCR positive plasma units had been transfused to 12 further patients. HCV infections were diagnosed in seven of these recipients, three patients had already been deceased. One of the remaining two recipients was already HCV positive prior to transfusion, in the other patient, no HCV infection was detectable. This patient had received three units of an "early" plasma donation , which was tested negative by PCR in one laboratory, but positive in the other. The subsequent, clinically infectious donation had the same discrepant PCR results. Thus, eight cases of HCV transmission were revealed and classified as "certain" with regard to causality, also due to an identical HCV genotype, i.e. 3e. Some of these infections would have been prevented by application of a different anti-HCV assay. The assay used in the respective plasmapheresis station was in-sensitive in this individual case for more than 400 days after the first PCR positive donation. This caused the release of the above mentioned infectious units. Upon re-testing the backups, three of four other anti-HCV assays revealed a positive result already 104 days after the first PCR-positive donation. The donor had increased ALAT levels (> 23 IU/L) at nine of 28 donations, two of these were higher than 2.5 times the upper normal limit, and two were higher than 68 IU/L, which is the cut-off value for male blood donors in Germany. The results of these (look-back) studies arouse several queries, i.e. differences in the diagnostic sensitivity between current anti-HCV and PCR tests, the accuracy of risk-estimates (especially when based on hemovigilance studies for Q-FFP), the value of ALAT testing, and currently practised release algorithms for Q-FFP.