Existing chronic kidney disease (CKD) is among the most potent predictors of postoperative acute kidney injury (AKI). Here we quantified this risk in a multicenter, observational study of 9425 patients who survived to hospital discharge after major surgery. CKD was defined as a baseline estimated glomerular filtration rate <45 ml/min per 1.73 m(2). AKI was stratified according to the maximum simplified RIFLE classification at hospitalization and unresolved AKI defined as a persistent increase in serum creatinine of more than half above the baseline or the need for dialysis at discharge. A Cox proportional hazard model showed that patients with AKI-on-CKD during hospitalization had significantly worse long-term survival over a median follow-up of 4.8 years (hazard ratio, 1.7) [corrected] than patients with AKI but without CKD.The incidence of long-term dialysis was 22.4 and 0.17 per 100 person-years among patients with and without existing CKD, respectively. The adjusted hazard ratio for long-term dialysis in patients with AKI-on-CKD was 19.8 compared to patients who developed AKI without existing CKD. Furthermore, AKI-on-CKD but without kidney recovery at discharge had a worse outcome (hazard ratios of 4.6 and 213, respectively) for mortality and long-term dialysis as compared to patients without CKD or AKI. Thus, in a large cohort of postoperative patients who developed AKI, those with existing CKD were at higher risk for long-term mortality and dialysis after hospital discharge than those without. These outcomes were significantly worse in those with unresolved AKI at discharge.