The current pandemic has generated the search for new reliable and economic alternatives for the detection of SARS-CoV-2, which produces the COVID-19 disease, one of the recommendations by the World Health Organization, is the detection of the virus by RT-qPCR methods from upper respiratory tract samples. The discomfort of the pharyngeal nasopharyngeal swab described by patients, the requirement of trained personnel, and the generation of aerosols, are factors that increase the risk of infections in this type of intake. It is known that the main means of transmission of SARS-CoV-2 is through aerosols or small droplets, which is why saliva is important as a relevant means of detecting COVID-19. In this study, a modified method based on SARS-CoV-2 RNA release from saliva is described, avoiding the isolation and purification of the genetic material and its quantification of viral copies; the results are compared with paired pharyngeal/nasopharyngeal swab samples (EF/EN). Results showed good agreement in saliva samples compared to EF/EN samples. On average, a sensitivity for virus detection of 80% was demonstrated in saliva samples competing with EF/EN samples. The use of saliva is a reliable alternative for the detection of SARS-CoV-2 by means of RT-PCR in the first days of infection, having important advantages over the conventional method. Saliva still needs to be studied completely to evaluate the detection capacity of the SARS-CoV-2 nucleic acid, however, the described process is viable, due to the decrease in materials and supplies, process times, the increment in the sampling and improvement of laboratory performance.