Abstract Chronic kidney disease is escalating public health concern in South Africa, however; data concerning its burden and risk factors are scant among critically ill patients due to lack of a national renal registry. The aim was to identify CKD predictors amongst critically ill patient admitted at the multi-disciplinary ICU in South Africa with emphasis on the contribution of the patient's creatinine level. A binary logistic regression analysis of 848 ICU admissions from the prospectively collected, publicly available, cohort dataset (LIVAKI), comprising data from 2017, was used to ascertain contribution of patient's peak creatinine value, age, gender and race to the burden of CKD. The analysis revealed that both age and patient creatinine level were statistically significant predictors of CKD (p < 0.001), while gender and race showed no significant effect. The likelihood ratio chi-square of 88.02 (p < 0.001) indicated that the model as a whole fit significantly. The results found are concordant with the growing evidence from sub-Saharan Africa showing that major contributors to burden of CKD in this region are chronic diseases like hypertension and diabetes mellitus; infectious diseases namely HIV and TB add to the problem This points to the role for systematic CKD screening in acutely ill patients and that peak creatinine level upon ICU admission could serve as a prognostic indicator for CKD. The study provided additional evidence to the limited body of literature describing CKD in the acutely ill South African setting, suggesting that proactive strategies to enhance screening, diagnosis and integration into the national health service delivery should be considered. Keywords: chronic kidney disease, creatinine, critical illness, intensive care unit, South Africa, acute kidney injury, risk factors, logistic regression