A modern binary executable is made up of various networks. This study is an empirical 1 analysis of the networks composing malicious binaries from multiple samples and quantifies their 2 structural composition with network measurements. We demonstrate the presence of Scale-Free 3 properties for data dependency and control flow graphs, and show that data dependency graphs 4 have both Scale-Free and Small-World properties. We show that program data dependency graphs 5 have a degree correlation that is disassortative, and that control flow graphs have a neutral degree 6 assortativity. These network measurements provide a set of features for further classification tasks to 7 identify patterns of malicious programs.