Abstract Aims The 2021 Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation combining creatinine and cystatin C provides a better estimation of glomerular filtration rate (GFR) compared to the creatinine-only equation. Methods and results CKD-EPI creatinine-cystatin C equation (creatinine-cystatin) was compared to creatinine-only (creatinine) equation in a subpopulation of Prospective comparison of ARNI with ACEI to Determine Impact on Global Mortality and morbidity in Heart Failure (PARADIGM-HF). Patients were categorized according to difference in eGFR using the two equations: Group 1 (<−10 mL/min/1.73 m2, i.e. creatinine-cystatin more than 10 mL/min lower than creatinine), Group 2 (>−10 and <10 mL/min/1.73 m2), and Group 3 (>10 mL/min/1.73 m2, i.e. creatinine-cystatin more than 10 mL/min higher than creatinine). Cystatin C and creatinine were available in 1966 patients at randomization. Median (interquartile range) eGFR difference was −0.7 (−6.4–4.8) mL/min/1.73 m2. Compared to creatinine, creatinine-cystatin led to a substantial reclassification of chronic kidney disease stages. Overall, 212 (11%) and 355 (18%) patients were reallocated to a better and worse eGFR category, respectively. Compared to patients in Group 2, those in Group 1 (lower eGFR with creatinine-cystatin) had higher mortality and those in Group 3 (higher eGFR with creatinine-cystatin) had lower mortality. Increasing difference in eGFR (due to lower eGFR with creatinine-cystatin compared to creatinine) was associated with increasing elevation of biomarkers (including N-terminal pro-B-type natriuretic peptide and troponin) and worsening Kansas City Cardiomyopathy Questionnaire clinical summary score. The reason why the equations diverged with increasing severity of heart failure was that creatinine did not rise as steeply as cystatin C. Conclusion The CKD-EPI creatinine-only equation may overestimate GFR in sicker patients. Clinical Trial Registration URL: https://www.clinicaltrials.gov; Unique Identifier: NCT01035255.