Prediction markets (PM) have drawn considerable attention in recent years as a tool for forecasting events. Studies surveying and examining relevant the trends of PM using traditional approaches have been reported in the literature. However, research using meta-analysis to review Prediction markets systems is very limited in Management Information System (MIS). This paper aimed to fill this gap by using Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) method to study Prediction markets trends over the past decades. Our results are as follows. First, we find that shows that more than 64% of academic studies on Prediction markets are published in top journals such as Journal of the Association for Information Systems, Journal of Consumer Research and Information Systems Research. Second, we showed that Prediction markets applications can be can be divided into two groups: internal use PMS and general public usage. Finally, our significant meta-analysis result show that on average prediction markets is 79% more accurate than alternative forecast methods based.