Determination of retroviral load is an important tool in the investigation of the success of therapeutic or vaccination trials in patients infected with lentiviruses such as HIV, or with their simian (SIV) or feline (FIV) counterparts. We have developed an one-tube quantitative reverse transcriptase-polymerase chain reaction (RT-PCR) assay based on the ABI Prism 7700 Sequence Detection System (TaqMan) to quantify the viral load of FIV-infected cats. Two different primer/probe systems were designed to detect a broad range of clade A FIV isolates. Both systems are characterized by excellent reproducibility, high sensitivity, and a wide range of quantification. As a consequence of this improved precision in the quantitative RT-PCR, preassay variations have greater impact on the accuracy of the viral load estimation. To compensate for these variations, we improved the assay and developed a multiplex real-time RT-PCR, which allows simultaneous calculation of the viral copy number and the individual recovery rate in an one-tube reaction. This enables the rapid and accurate calculation of copy number independent of preassay variations. In further studies, two additional real-time RT-PCR assays were designed and used to investigate the influence of sequence variations in the binding regions for either the primers or probe. Sequence mismatches in this region had a significant effect (up to 4 logarithmic decades) on reaction efficiency. In view of the inherent variability of retroviral sequences, these results underline the necessity to check reaction efficiencies before determining viral load.