Objective To evaluate the effect of early non-liquid diet for patients with mild acute pancreatitis by using Meta-analysis. Methods The randomized controlled trials(RCT) on early non-liquid diet intake for patients with mild acute pancreatitis were collected by computer searching the database of China National Knowledge Infrastructure(CNKI),Wanfang Periodical Database,CBM,Elsevier Scien Direct, cochrane library and PubMed from inception to February 2017, and analyzed by RevMan 5.2 software. Results Five RCTs were included in the study.The Results of Meta-analysis revealed that both the length of hospitalization after eating [MD=-1.48,95%CI-2.50~-0.47, P=0.004] and the total length of hospitalization[MD=-1.99,95%CI-(0.99-2.99),P=0.0001]in liquid diet group were less than that of non-liquid diet group,and the differences were statistically significant.There was no significant difference in the rate of the pain recurrence[RR=1.21,95%CI 0.77-1.91,P=0.40]and the rate of the elimination of oral feeding[RR=1.02,95%CI 0.45-2.30,P=0.96]because of pain,nausea or vomiting after refeeding,although all that rate in non-liquid diet was not lower than that of liquid diet group. Conclusions compared with liquid diet,early non-liquid diet intake for mild acute pancreatitis can shorten their hospitalization time, and it cannot be concluded that the non-liquid diet will increase the pain recurrence rate, and most patients have a better tolerance to non-liquid diet,and non-liquid diet can promote rehabilitation.