Although non-exercise estimated cardiorespiratory fitness (eCRF) algorithms correlate well with measured CRF, their prognostic value in high-risk populations remains unclear. We investigated the associations between eCRF and mortality risk in adults with diabetic kidney disease (DKD), a vulnerable population with elevated mortality risk.Data from adults with DKD were derived from the 1999-2018 National Health and Nutrition Examination Survey. The primary exposure was eCRF, which was calculated from validated algorithm incorporating age, sex, body mass index, waist circumference, resting heart rate, physical activity, and smoking status. The primary outcomes were all-cause and cardiovascular disease (CVD) mortality. Hazard ratios (HRs) and 95 % confidence intervals (CIs) were calculated from Cox proportional hazards models for the associations between eCRF and mortality risk.Among 2045 participants (mean age 61.65 years, 52.10 % men) included, 804 (35.84 %) deaths occurred during a median follow-up of 84 months, of which 296 were from CVD causes. Compared with the highest eCRF quartile, the HRs (95 % CIs) of the lowest quartile were 1.75 (1.27-2.40) for all-cause mortality and 2.29 (1.32-3.99) for CVD mortality. Restricted cubic spline curves demonstrated inversely linear relationships between eCRF and all-cause and CVD mortality. Associations between eCRF and all-cause mortality were stronger in female participants, those with overweight or obesity, and those without prior CVD.Lower eCRF was independently associated with elevated risks of all-cause and CVD mortality in individuals with DKD, supporting the utility of eCRF as a practical prognostic tool in clinical management of DKD.Copyright © 2025 The Authors. Published by Elsevier Inc. All rights reserved.