Severe acute respiratory syndrome coronavirus 2(SARS-CoV-2)and coronavirus disease 2019(COVID-19)continue to impact countries worldwide.At present,inadequate diagnosis and unreliable evaluation systems hinder the implementation and development of effective prevention and treatment strategies.Here,we conducted a horizontal and longitudinal study comparing the detection rates of SARS-CoV-2 nucleic acid in different types of samples collected from COVID-19 patients and SARS-CoV-2-infected monkeys.We also detected anti-SARS-CoV-2 antibodies in the above clinical and animal model samples to identify a reliable approach for the accurate diagnosis of SARS-CoV-2 infection.Results showed that,regardless of clinical symptoms,the highest detection levels of viral nucleic acid were found in sputum and tracheal brush samples,resulting in a high and stable diagnosis rate.Anti-SARS-CoV-2 immunoglobulin M(IgM)and G(IgG)antibodies were not detected in 6.90％of COVID-19 patients.Furthermore,integration of nucleic acid detection results from the various sample types did not improve the diagnosis rate.Moreover,dynamic changes in SARS-CoV-2 viral load were more obvious in sputum and tracheal brushes than in nasal and throat swabs.Thus,SARS-CoV-2 nucleic acid detection in sputum and tracheal brushes was the least affected by infection route,disease progression,and individual differences.Therefore,SARS-CoV-2 nucleic acid detection using lower respiratory tract samples alone is reliable for COVID-19 diagnosis and study.