Abstract Background and Aims Exercise reduced risk of diabetes and obesity. However, it remains unclear whether the pattern changes of exercise are increased risk of end-stage renal disease in new-onset diabetes mellitus (DM) patients. Method We assessed diabetes development in 2015 in 21,122,422 non-diabetic Koreans enrolled in Korean National Health Screening (KNHS) from 2009 to 2012 using International Classification of Diseases 10th (ICD-10) code and anti-diabetic medication prescription. Renal outcome (End-stage renal disease) was defined using V-code. The risk of ESRD evaluated in never exercise (none), after new-onset DM (post only), before new-onset DM and no exercise after new-onset DM (pre-only) and persistent exercise group (persistent group). Results During median 4.8-year follow-up, 275 (0.21%) of ESRD was occurred. After fully adjusted, the hazard ratio (HR) for ESRD was lowest in persistent exercise group 0.521 [0.302 to 0.899]. Post only exercise 0.674 [0.461 to 0.985] group showed better renal outcome compared to pre only exercise group 0.791 [0.535 to 1.169] or never exercise group [reference, 1.0]. Subgroup analysis for age, sex, history of chronic kidney disease, hypertension showed similar results. Even light exercise such as walking tended to reduce the risk of renal outcome 0.817 [0.634 to 1.037]. The higher 0.627 [0.441 to 0.892] or moderate 0.7 [0.514 to 0.953] strength of exercise type showed the lower risk of renal outcome in new-onset DM patients. Conclusion This study provides compelling evidence that exercise, particularly when maintained over time, is associated with a substantial reduction in the risk of ESRD in individuals with new-onset DM. These findings underscore the importance of incorporating regular physical activity as a crucial component of comprehensive health management strategies for individuals with diabetes, offering both clinicians and patients valuable insights for optimizing renal outcomes.