Summary What is known and objective Cystatin C (Cys‐C) is a useful diagnostic marker for early renal dysfunction, but has the disadvantage of giving false‐positive results when corticosteroids are administered. In this study, we aimed to evaluate the dose‐dependent effect of corticosteroids on the divergence between estimated glomerular filtration rates based on Cys‐C (eGFRcys) and creatinine (eGFRcreat) and calculate the cut‐off value of corticosteroid dose having an impact on eGFRcys/eGFRcreat ratio. Methods This retrospective study included 305 patients (1318 therapies) treated with oral or injectable corticosteroids between June 2014 and May 2018, who did not meet the exclusion criteria. All corticosteroid doses were converted to prednisolone equivalent. Results Steroid dose correlated significantly with eGFRcys/eGFRcreat ratio for all corticosteroids and for prednisolone (rs = −.150 and −.273, respectively), whereas no correlation was observed for methylprednisolone and hydrocortisone. The cut‐off value of prednisolone dose for eGFRcys/eGFRcreat ratio < 0.79 was 0.170 mg/kg/day, with 62.4% sensitivity and 84.7% specificity. The correlation coefficient (rs = −.434) between prednisolone dose and eGFRcys/eGFRcreat ratio for doses of 0.170 mg/kg/day and higher was markedly larger compared with all corticosteroids. What is new and conclusion These findings suggest that results should be interpreted with caution when using eGFRcys as renal function marker in patients treated with prednisolone at doses of 0.170 mg/kg/day and higher.