In this chapter, we present and interpret some operations on biological networks that can easily performed with NeAT, a set of Web tools aimed at studying biological networks (or graphs) and classifications. These approaches are of particular interest for biologists and scientists who need to assess the reliability of new datasets (either experimental or predicted) by comparing them to established references. Firstly, we describe the steps that will allow a nonspecialist user to compare two networks to compute their union and the statistical significance of their intersection. Next, we show how to map functional classes (e.g., GO categories, sets of regulons or complexes) onto a biological network. A third protocol explains how to compare two sets of functional classes, e.g., to assess statistically the biological relevance of some computationally returned groups of genes (clustering). The metrics as well as the results obtained by following the different protocols are extensively described and explained. NeAT is available at the following URL: http://rsat.bigre.ulb.ac.be/rsat/index_neat.html.