Objectives: Procalcitonin (PCT), a well-recognized biomarker for diagnosing infection and guiding antibiotics therapy, has been proposed to have prognostic value for predicting mortality and the occurrence of acute kidney injury (AKI) in septic patients. In this study, we investigated whether PCT predicts mortality and renal recovery in critically ill patients with sepsis-induced acute kidney injury (SIAKI) receiving continuous renal replacement therapy (CRRT). Methods: We examined 649 patients with SIAKI who received CRRT in our intensive care unit. %PCT decrease was calculated using the following formula: ([PCT level at CRRT initiation – PCT level at 72h after CRRT initiation]/ PCT level at CRRT initiation) × 100.  Results: Of the 649 patients, 332 expired within 28 days after CRRT initiation. Multivariable Cox regression analysis demonstrated that a single PCT level at CRRT initiation was not associated with mortality among all patients. However, %PCT decrease was independently associated with the mortality (per 10% decrease, hazard ratio [HR] = 0.87; P < 0.001). Among the survivors (n = 332), %PCT decrease, not a single PCT level at CRRT initiation, was independently associated with renal recovery from dialysis (per 10% decrease, HR = 1.28; P < 0.001). The best cutoff value of %PCT decrease for predicting survival was > 31% (area under the curve [AUC] = 0.802; P < 0.001). Among survivors, the best cutoff value for predicting the renal recovery from dialysis was > 69% (AUC = 0.825; P < 0.001). Conclusions: This study demonstrated that PCT decrease was independently associated with mortality and renal recovery from dialysis in critically ill patients with SIAKI receiving CRRT. PCT decrease, not a single PCT level, could be a valuable tool for predicting prognosis in these patients.