Clinical guidelines recommend continuation of sodium-glucose cotransporter-2 inhibitor (SGLT2i) treatment when renal function deteriorates among patients with type 2 diabetes. However, the recommendation is not currently supported by research specifically designed to compare continuation versus discontinuation of SGLT2is for patients who have received the treatment for a while before experiencing renal function deterioration. Using linked Taiwanese databases with claims and clinical data and a target trial emulation framework, we conducted a nationwide cohort study to investigate the association between SGLT2i continuation and major cardiorenal outcomes after renal function declined. We identified patients with type 2 diabetes who received SGLT2is during 2016-2020, who either continued or discontinued SGLT2is within 180 days after their eGFR declined below 45 mL/min/1.73 m. The study outcomes were myocardial infarction, stroke, heart failure, acute kidney injury, and all-cause mortality. We estimated the on-treatment and intention-to-treat hazard ratios (HRs) and 95% confidence intervals (CIs) comparing SGLT2i continuation versus SGLT2i discontinuation, adjusting for baseline and time-varying covariates using inverse probability weighting. Among 11,467 eligible patients, the on-treatment HR associated with SGLT2i continuation was 0.83 (95% CI, 0.59-1.18) for myocardial infarction, 0.74 (0.56-0.95) for stroke, 0.58 (0.49-0.69) for heart failure, 0.53 (0.45-0.63) for acute kidney injury, and 0.51 (0.42-0.61) for all-cause mortality. The results did not change meaningfully in the intention-to-treat analysis or within pre-specified patient subgroups, including patients with eGFR below 30 mL/min/1.73 m or prior acute kidney injury. Our findings provide scientific evidence to support continuing SGLT2i use when renal function declines during the SGLT2i treatment course.