Background: Changes in postoperative serum creatinine levels have been used to define acute renal injury in patients undergoing cardiac surgery with cardiopulmonary bypass. It remains unclear, however, whether subclinical increases in serum creatinine that do not meet current Acute Kidney Injury Network or RIFLE (risk, injury, failure, loss, and end-stage kidney disease) criteria for acute renal injury are predictive of mortality after cardiac surgery. Methods: Multivariate logistic regression was performed in a retrospective cohort of 3914 consecutive patients undergoing primary, isolated coronary artery bypass grafting with cardiopulmonary bypass to determine whether postoperative serum creatinine change independently predicts 30-day all-cause mortality in patients with normal renal function and with varying levels of preoperative renal insufficiency. To control further for selection bias, multivariate logistic regression was performed on a propensity-matched cohort (n = 2042) to determine whether subclinical increases in serum creatinine predict mortality. Results: Negative change in serum creatinine was associated with reduced 30-day all-cause mortality. Even subclinical increases in serum creatinine were associated with increased mortality relative to patients with negative changes in serum creatinine (odds ratio, 3.93; 95% confidence interval, 1.68-9.22; P < .01). After propensity matching, subclinical increases in serum creatinine were still associated with increased mortality (odds ratio, 4.13; 95% confidence interval, 1.37-12.45; P = .01). Conclusions: Subclinical increases in serum creatinine that do not meet acute renal injury criteria are independently associated with 30-day all-cause mortality in patients with normal renal function or preoperative renal insufficiency undergoing coronary artery bypass grafting.