Abstract Context Patients with diabetes are at a higher risk of pneumonia and pneumonia mortality. Sodium glucose co-transporter 2 inhibitors (SGLT2is), the latest class of glucose-lowering agents, were shown to reduce the risk of pneumonia in clinical trials. However, the real-world effectiveness of SGLT2is on the risk of pneumonia is largely unknown. Objective To investigate the associations between SGLT2is use and the risk of pneumonia and pneumonia mortality compared with dipeptidyl peptidase-4 inhibitors (DPP4is) using an electronic medical database in Hong Kong. Design A retrospective cohort study. The “prevalent new-user” design was adopted to account for the previous exposure to the study drugs being compared. Propensity score (PS) matching (1:4) was used to balance the baseline characteristics of the 2 groups. Setting and participants Electronic health data of type 2 diabetes patients using SGLT2is and DPP4is between 2015 and 2018 was collected from the Clinical Data Analysis and Reporting System. Main Outcome Measures Pneumonia incidence and mortality. Results The PS-matched cohort consisted of 6664 users of SGLT2is and 26 656 users of DPP4is, with a mean follow-up of 3.8 years. Poisson regression showed that SGLT2is use was associated with lower risk of pneumonia compared with DPP4is with an absolute rate difference of 4.05 per 1000 person-years (95% CI, 2.61-5.51). The corresponding incidence rate ratio was 0.71 (95% CI, 0.62-0.81). Similar reduction in risk of pneumonia death was observed (hazard ratio 0.57; 95% CI, 0.42-0.77). Conclusion Compared with DPP4is, SGLT2is use was associated with a reduced risk of pneumonia and pneumonia mortality in a real-world setting.