Early recognition of acute kidney injury (AKI) in the intensive care unit (ICU) remains a critical problem, with a rising incidence and a high mortality rate. As a consequence, the actual lack of an early and effective biomarker results in a significant delay in initiating appropriate therapy. The accurate diagnosis of AKI is especially problematic in critically-ill patients, in whom we know that renal function is in an unsteady state; therefore the validity of creatinine- based baseline assessment measures is reduced. Because the rationale for assessing AKI markers in critically-ill patients is strong at the present time, researchers are stimulated to establish a multidimensional AKI classification system. This system should in essence grade AKI severity. The most widely referenced classification is the RIFLE system. Thus, early recognition of AKI, well before changes in serum creatinine occur, has come under intensive research, because it is evidenced that even small increases in serum creatinine are associated with an increase in patient mortality. The development of a biomarker kit in which several early markers with different characteristics are combined, is essential. Multi-centre, randomized studies indicate a potential for early biomarkers able to diagnose AKI 48 hours before creatinine changes. In conclusion, time has come to leave serum creatinine behind as a marker of renal function in patients with AKI on the ICU. Only then will we be able to offer early goal-directed therapy for the kidney in the ICU setting.