Acute pancreatitis (AP) continues to be a clinical challenge. The mortality of patients with AP with adverse outcomes like organ failure and infected necrosis can be as high as 43 %. Highly accurate predictors of adverse outcomes are necessary to identify the high-risk patients so that they can be meticulously monitored and managed. However, there are no ideal predictors till date. Over the past several years, a number of single- and multi-parameter predictors have been identified and tested for prediction of adverse outcomes in AP. Out of the different tools tested, blood urea nitrogen and the harmless acute pancreatitis score appears to be useful and feasible in the management of AP under Indian conditions. Other single-parameter predictors like serum creatinine, hematocrit, erythrocyte sedimentation rate, C-reactive protein, and D-dimer need to be put to further tests in high-quality prospective studies with large sample size at the community level. Multi-parameter prediction tools like the bedside index of severity of acute pancreatitis may not be appealing in day-to-day clinical practice.