We compare the accuracy of the Normalized Compress Distance (NCD) technique for building phylogenetic trees from molecular data with the more widely used techniques of Neighbor joining with p-value and Bayesian inference. The NCD algorithm is a distance-based alignment-free method and as such takes unaligned sequence data as input and outputs a distance matrix. The matrix is given as input to FastME, which outputs a tree in Newick format. Every step of this process requires minimal processing and as such provides fast execution times along with minimal restrictions on input data.