Abstract: This study presents the performance evaluation of Relevancy Measure based MapReducer (RM-MapReducer) in supporting web page categorization with different datasets between 200 and 400 web pages as test data. Experiments were performed using datasets of WebKB and ODP with the real time crawler dataset. It was observed that the real time crawler dataset provides better results compared to the other dataset of WebKB and ODP using MapReduce programming model.