Acute pancreatitis (AP) is an inflammatory disease of the pancreas that can lead to gland necrosis, and end-organ failure in up to 20% of cases. Given the prognostic implication of severity, early prediction of severe disease thus has been a subject of intense research. Over the last several decades, over a dozen prediction tools have been developed using various clinical parameters including examination findings, radiological features, laboratory values, and host-related characteristics. Existing prediction tools attempt to assess at least one or more of the following: presence of volume deficit; severity of host inflammatory response; age and comorbidity burden; risk of fat saponification; and early organ dysfunction. In this chapter, the authors discuss each of these pathophysiological mechanisms, and prediction tools that attempt to assess these early in the course of the disease. Advantages, disadvantages, and future direction are also discussed.