Background/Aims: Type 2 diabetes mellitus (T2DM) is an extremely prevalent disease with multisystem complications. We aim to compare the effects of two common glucose lowering medications; sodium glucose co-transporter 2 inhibitors (SGLT2I) and dipeptidyl peptidase-4 inhibitors (DPP4I), on the incidence of diabetic retinopathy and cataracts in T2DM patients in Hong Kong. Methods: Retrospective population-based cohort study of T2DM patients treated with SGLT2I or DPP4I between 1st January 2015 and 31st December 2020. Propensity score matching (1:1 ratio) between SGLT2I and DPP4I users was performed on demographics, past co-morbidities, number of prior hospitalizations, duration from T2DM diagnosis to intial drug exposure, non-SGLT2I/DPP4I medications (including other anti-diabetes drugs), abbreviated modification of diet in renal disease, HbA1c, fasting glucose, and their time-weighted means. Sensitivity analysis using a one-year lag time and competing risk analyses using cause-specific and sub-distribution hazard models were conducted. Results: This study cohort included 26 165 SGLT2I and 42 796 DPP4I users (total: N=68 961 patients; 56.43% males, median age: 62.0 years old (standard deviation (SD): 12.8)). Over a median follow-up of 5.56 years (IQR: 5.24-5.80) and after propensity score matching (SGLT2I: N=26 165; DPP4I: N=26 165), SGLT2I users had lower incidences of cataract (4.54% vs. 6.64%%, standardised mean difference [SMD]=0.09) and diabetic retinopathy (3.65 vs. 6.19, SMD=0.12) compared to DPP4I users. SGLT2I use was associated with lower risks of new onset cataract (HR: 0.67, 95% CI: [0.62-0.72] P<0.0001) and diabetic retinopathy (hazard ratio [HR]: 0.57, 95% confidence interval [CI]: [0.53-0.62], P<0.0001). These associations remained significant on multivariable Cox regression ;cataract: HR: 0.69, 95% CI: 0.64-0.75 (P<0.0001); diabetic retinopathy: HR: 0.68, 95% CI: 0.63-0.75 (P<0.0001). Conclusions: Amongst T2DM patients in Hong Kong, SGLT2I use was associated with lower risks of new onset cataract or diabetic retinopathy compared to DPP4I use.