Objective: To invesigate the protective effect of curcumin for severe acute pancreatits and its possible mechanism. Methods: Fifty-four SD rats were equally randomized into the sham operation group, SAP model group (SAP group) and curcumin pretreatment + SAP group (curcumin pretreatment group). The SAP model was induced by retrograde cholangiopancreatic duct injection of 5% sodium taurocholate. At 3, 6 and 12 hours after surgery, the ascites volume and the following observation parameters of each group were determined: these included the serum levels of amylase (AMY), IL-6 and TNF-α, and the pathological impairment and scoring of the pancreatic tissues as well as the expression of NF-κB p65 and TLR4 in the pancreatic tissues. Results: The ascites volume and serum levels of AMY, IL-6 and TNF-α in both the SAP group and curcumin pretreatment group were markedly increased at each observation time point compared with the sham operation group, especially at 6 and 12 hours, and the differences between them had statistical significance (all P<0.01). Except for the sham operation group, both the SAP group and curcumin pretreatment group exhibited the typical pathological changes of SAP, while the pathological changes in the curcumin pretreatment group at each time point were all markedly milder than those in the SAP group and the Schmidt scores at the 6- and 12-h time points of the curcumin pretreatment group were all significantly lower than those of the SAP group (both P<0.05). The expression levels of NF-κB p65 and TLR4 in the SAP group and curcumin pretreatment group were markedly higher than those of the sham operation group but both expressions of the curcumin pretreatment group were lower than those of the SAP group. Conclusion: Curcumin has protective effect for SAP, and the possible mechanism may be associated with its inhibition of the activation of TLR4 and NF-κB, thereby alleviating the inflammatory response and pancreatic injury.