Objective To investigate the association between the normalized serum creatinine-to-cystatin C ratio (NCCR) and new-onset stroke in middle-aged and elderly adults, and provide new evidence for stroke prevention and treatment. Methods Based on data from the China Health and Retirement Longitudinal Study (CHARLS), 10,526 participants who completed the surveys in 2015, 2018, and 2020 were included. Participants were grouped by NCCR quartiles. Cox proportional hazards regression models were used to analyze the relationship between NCCR and new-onset stroke. Restricted cubic splines were applied to explore the dose-response relationship. Subgroup analyses (by gender, age, smoking status, and drinking status) were conducted to verify subgroup differences in the association, and mediation analyses were performed to identify the mediating roles of cardiometabolic factors (blood pressure, blood glucose, blood lipids, and inflammatory markers). Results During follow-up, 721 participants (6.85%) developed new-onset stroke. Cox proportional hazards regression analysis showed that after adjusting for all confounding factors, compared with the Q1 group, the risk of stroke was reduced by 20% in the Q3 group (HR=0.80, 95%CI:0.65-0.99) and 37% in the Q4 group (HR=0.63, 95%CI: 0.49-0.80). Additionally, the risk of stroke decreased with increasing NCCR levels (P for non-linearity >0.05). Further mediation analysis revealed that systolic blood pressure and diastolic blood pressure were key mediators explaining the observed relationship between NCCR and stroke risk, with mediating proportions of 34.50% and 29.37%, respectively. Conclusion A high level of NCCR may reduce the risk of stroke onset, with both systolic blood pressure and diastolic blood pressure playing a key mediating role