Résumé : Objectives: Nasopharyngeal swab sampling for the detection of SARS-CoV-2 is a standard method for the diagnosis of COVID-19, but its collection usually causes discomfort in the patient and exposes healthcare workers to a higher risk. Saliva seems to be a good alternative to nasopharyngeal swabs, as it is non-invasive, reduces the risk of contamination of healthcare workers, and allows self-collection. This study aims to compare the ability to detect SARS-CoV-2 by RT-PCR in the same patient using saliva and nasopharyngeal swab samples to analyze the concordance of the results obtained between the two samples. Methods: Thirty saliva and nasopharyngeal swab samples from patients with COVID-19 symptoms who were admitted to the emergency department of the Viedma Clinical Hospital were taken in parallel. Both samples were analyzed by RT-PCR for the detection of SARS-CoV-2. The concordance of results was calculated using the Cohen's Kappa coefficient. Results: Our results show that there is good concordance (Kappa index 0.730; 95% CI: 0.486-0.974) between the two types of samples analyzed. Conclusions: Saliva seems to be a reliable and effective sample for the detection of SARS-CoV-2.