Three examples of human plasma-derived concentrates, intermediate-purity factors VIII and IX, and fibrinogen were spiked with tissue culture-grown human immunodeficiency virus type 1 (HIV-1) strain RF. All examples were freeze-dried and heated at 80 degrees C for 72 hours by using validated production process models. HIV-1 infectivity was measured by a syncytial infectivity assay in C8166 cells and then compared with levels determined by nested HIV polymerase chain reaction (PCR). The infectivity assay demonstrated a reduction index of at least 4.5 log10, while PCR showed an average 1.7 log10. Large amounts of HIV-1 RNA (10(5)) were still detectable by PCR in samples in which infectivity assays failed to detect any HIV-1. These data suggest that HIV-1 PCR levels do not parallel HIV-1 infectivity levels during virus-inactivation procedures involved in coagulation factor concentrate production. PCR was able to detect the RNA associated with inactivated HIV-1 particles in the factor concentrates, which allows the conclusion that PCR is not a useful test with which to monitor virus-inactivation procedures such as heating at 80 degrees C for 72 hours. This judgment contrasts with the more definite and sensitive role of PCR in diagnosing HIV-1 infection in patients in whom a positive HIV-1 PCR result correlates with active HIV-1 infection and with PCR's usefulness in monitoring virus removal.