Acute dysfunction of the kidneys is common in intensive care mandating accurate and timely diagnosis with major therapeutic and prognostic implications. The concept of acute kidney injury (AKI) based on serum creatinine levels or urine output has been used in several definitions of AKI. Creatinine is the golden standard of renal function measurement, but its relation to glomerular filtration rate is not reliable in critical illness. Cystatin C, another endogenous substance used for GFR assessment, has been reported to be better at assessing renal function in the ICU. Given its more advantageous profile to estimate GFR especially at higher mild to moderate decrease in GFR, identifying patients with AKI with cystatin C could be useful since this could alert the clinician earlier than with creatinine to consider actions for prevention or treatment of AKI. We hypothesized that serum/plasma cystatin C derived eGFR levels corresponding to current AKI stages would be a better predictor of outcome than creatinine-based AKI grading. Accordingly, we study a cystatin C based AKI grading and assess its performance in a large Swedish cohort or of critically ill patients.