RT-PCR is the main method for declaring that someone is COVID-19 infected or not, as well as having numerous other uses in molecular biology research and biological testing. Professor Stephen Bustin is a world expert on the technology, and the potential problems with using it to produce accurate and repeatable results. Although the coronavirus test is presented as a binary test, it is actually based on whether the production of DNA is detectable prior to an arbitrary number of PCR cycles. If there is variability in the quantification, then samples will be above or below the limit, when they should not be, resulting in false positives and negatives. David and Stephen walk through the steps, from the extraction of RNA from the original sample, the conversion of the RNA to complementary DNA, and duplication of DNA using PCR, and the optional step of sequencing. While this is dense technical information at times, it is presented logically, and the limitations of this method cannot be understood without taking the cover off the black box. We suggest not listening to this episode when you are trying to do anything else, but sit down in a quiet place so that you can concentrate fully.