Abstract: The main goal of this study is to analyse the morphological structure of the Turkish documents. The new proposed method consists of lossless compressing the monograms, digrams, trigrams, roots-stems and suffixes individually using a statistical approach. In this work 1g, 2 g, 3 g, root-system and suffix frequencies have been computed for Turkish language. A tuned template has been prepared for each group. The compression of Turkish documents has been performed by using the static Huffman Coding for Word Based Dynamic Huffman has been measured.