Objective: We aimed to analyse the positivity rate and cycle threshold values indicating viral loads for SARS CoV-2 among differentrespiratory specimens and also to evaluate the diagnostic efficacy of saliva samples.Materials and Methods: We included combined oropharyngeal and nasopharyngeal swab (cONS), sputum, and tracheal aspirate(TA) specimens of patients. Unpreserved saliva samples were collected prospectively from hospitalized patients within 72 hours ofadmission. SARS CoV-2 RNA was extracted by using Bio-Speedy viral nucleic acid buffer than RT-PCR was performed with BioSpeedy COVID-19 qPCR detection kit.Results: Retrospective evaluation revealed SARS CoV-2 RNA in 19.66% of cONS (n: 5819), 30.77% of sputum (n: 39), 29.41% of TAsamples (n: 34) from 4812 patients. In the majority (86.72%) of the samples, the first cONS sample was positive. Consecutive cONSand sputum/TA samples were investigated in 52 patients of whom 11 were positive with either of these samples. Saliva positivity wasdetected in 60% of cONS positive (n: 20) and 30% of cONS negative (n: 12) patients.Conclusion: Although, cONS samples show the greatest diagnostic guidance, repeated sampling from multiple sites of the respiratorytract increases the possibility of COVID-19 diagnosis. Saliva samples might be considered as an alternative specimen.