COVID-19 is an example of a big challenge for diagnostics. In the case of many diseases and syndromes, there is a possibility to repeat the tests, and the time allocated for diagnosis is counted even in weeks, not in hours, as in COVID-19. The situation related to COVID-19, therefore leads to a large number of tensions, both among diagnostic physicians and patients. Especially that in this case, based on a positive result, the decision is being made not only about the choice of a possible therapy, but also about quarantine. This article refers to the predictive value of tests used in the diagnosis of COVID-19 (mainly RT-PCR), focusing on the concept of a false positive result. The article demonstrates that this is a relative term, depending on the question posed, the purpose of the diagnosis (selection of therapy, quarantine decision, disease detection), or even the variant of the SARS-CoV-2 virus. Effective vaccination against COVID-19, the increasing percentage of convalescent patients, and the emergence of the SARS-CoV-2 omicron variant minimalize all these problems. However, shifting away from widely used screening tests, should not cause abandoning COVID-19 testing completely. Part of the population may still die from the disease and effective targeted therapy already exists. In order to personalize this therapy, a test such as RT-PCR must be used.