We compared the performance of a laboratory-developed 5'-nuclease real-time polymerase chain reaction assay and a commercial assay (LightCycler, Roche Diagnostics, Indianapolis, IN) for quantification of Epstein-Barr virus (EBV) DNA. Using standards provided by the manufacturer, the LightCycler assay was linear from 100 to 1 million copies per reaction. Based on dilution of a plasmid containing the amplicon, the laboratory-developed assay was linear from 22 to 45 million copies per reaction. Both assays detected 0.5 copies of genomic EBV DNA per reaction; both showed good reproducibility with coefficients of variation from 0.3% to 2.4% for the LightCycler and 1.8% to 5.1% for the laboratory-developed assay. For 31 whole blood specimens with measurable values in both assays, the viral load values obtained with the LightCycler averaged 2.3-fold higher than those obtained with the laboratory-developed assay. Testing 30 matched whole blood and plasma samples in the laboratory-developed assay showed whole blood viral load values averaged 10-fold higher than those for plasma. The LightCycler and laboratory-developed assays are sensitive and reproducible with broad linear ranges. Further clinical evaluation is needed to determine the viral load cutoff that is predictive of posttransplantation lymphoproliferative disorders.