Objectives: Evaluate the risk factors of prolonged SARS-CoV-2 virus shedding and the impact of arbidol treatment on SARS-CoV-2 virus shedding. Methods: Data were retrospective collected from adults hospitalized with COVID-19 in Wuhan Union Hospital. We described the clinical features and SARS-CoV-2 RNA shedding of patients with COVID-19 and evaluated factors associated with prolonged virus shedding by multivariate regression analysis. Results: Among 238 patients, the median age was 55.5 years, 57.1% were female, 92.9% (221/238) used arbidol, 58.4% (139/238) used arbidol combination with interferon. The median time from illness onset to start arbidol was 8 days (IQR, 5-14 days) and the median duration of SARS-CoV-2 virus shedding was 23 days (IQR, 17.8-30 days). SARS-CoV-2 RNA clearance was significantly delayed in patients who received arbidol >7 days after illness onset, compared with those in whom arbidol treatment was started less than or equal to7 days after illness onset (HR, 1.738 [95% CI, 1.339-2.257], P < .001). Multivariate regression analysis revealed that prolonged viral shedding was significantly associated with initiation arbidol more than seven days after symptom onset (OR 2.078, 95% CI [1.114-3.876], P .004), more than 7 days from onset of symptoms to first medical visitation (OR 3.321, 95% CI[1.559-7.073], P .002), illness onset before Jan.31, 2020 (OR 3.223, 95% CI[1.450-7.163], P .021). Arbidol combination with interferon was also significantly associated with shorter virus shedding (OR .402, 95% CI[.206-.787], P .008). Conclusions: Early initiation of arbidol and arbidol combination with interferon as well as consulting doctor timely after illness onset were helpful for SARS-CoV-2 clearance.