MapReduce is a highly efficient distributed and parallel computing framework, allowing users to readily manage large clusters in parallel computing. For Big data search problem in the distributed computing environment based on MapReduce architecture, in this paper we propose an Ant colony parallel search algorithm (ACPSMR) for Big data. It take advantage of the group intelligence of ant colony algorithm for global parallel search heuristic scheduling capabilities to solve problem of multi-task parallel batch scheduling with low efficiency in the MapReduce. And we extended HDFS design in MapReduce architecture, which make it to achieve effective integration with MapReduce. Then the algorithm can make the best of the scalability, high parallelism of MapReduce. The simulation experiment result shows that, the new algorithm can take advantages of cloud computing to get good efficiency when mining Big data.