During the COVID-19 pandemic, a certified laboratory of Tamaulipas, Mexico has processed over 100,000 samples of COVID-19 suspected patients, working a minimum of 100 tests daily. Thus, it would be beneficial for such certified laboratories nationwide to reduce the time and cost involved in performing the diagnosis of COVID-19, from sample collection, transportation to local lab, processing of samples, and data acquisition. Here, 30 nasopharyngeal swab and saliva samples from the same COVID-19 individuals were assessed by a standard nucleic acid extraction protocol, including protein lysis with proteinase K followed by binding to column, washing, and elution, and by the SalivaDirect protocol based on protein lysis, skipping the other steps to reduce processing time and costs. The genomic RNA was amplified using a SARS-CoV-2 Real-Time PCR kit. A variation (P > 0.05) in the 95% CIs = 72.6%–96.7% was noted by using the SalivaDirect protocol and saliva samples (sensitivity of 88.2%) in comparison to those of standard protocol with oropharyngeal swab samples (95% CIs = 97.5%–100%; sensitivity of 100%) as reported elsewhere. However, when using nasopharyngeal swab samples in the SalivaDirect protocol (sensitivity of 93.6%; 95% CIs = 79.2%–99.2%), it was in concordance (P < 0.05) with those of the standard one. The logical explanation to this was that two samples with Ct values of 38, and 40 cycles for gene E produced two false negatives in the SalivaDirect protocol in relation to the standard one; thus, there was a reduction of the sensitivity of 6.4% in the overall assay performance.