HBase and MongoDB are non-relational databases where HBase is column-oriented, and MongoDB is documentoriented. MongoDB uses indexing to speed up the searching process, while HBase does not have secondary indexing, a lookup table was created to save an index of searching. This research used a prototype developed in a similar data model design of the HBase and MongoDB database to compare the response time performance in searching operation using keyword frequency criteria in the document management system. The results show that the primary indexes of HBase is faster than secondary indexes of MongoDB. The time difference between HBase and MongoDB is almost a half.