Summary: Azvudine and nirmatrelvir-ritonavir are prioritized treatments for SARS-CoV-2 infection in China, but their effectiveness and safety in hospitalized patients with COVID-19 and pre-existing diabetes remains unknown. In this retrospective cohort study, we collected 32,864 hospitalized COVID-19 patients from nine hospitals, among which 636 azvudine recipients and 318 nirmatrelvir-ritonavir recipients were enrolled for final analysis after exclusion and propensity score matching. Kaplan-Meier and multivariate Cox regression analysis results showed that azvudine had a lower risk of all-cause death compared with nirmatrelvir-ritonavir for the treatment of patients with COVID-19 and pre-existing diabetes (log rank: p = 0.044; HR: 0.63; 95% CI: 0.431–0.934). No significant difference was found in composite disease progression between the two groups. Five sensitivity analyses verified the robustness of the results. Overall, azvudine may be comparable to nirmatrelvir-ritonavir in terms of both all-cause death and composite disease progression among hospitalized patients with COVID-19 and pre-existing diabetes, with acceptable adverse events.