Once three-dimensional reconstructions have been performed, most of the work required to access the information contained in the data set has to be done. Potential features of interest are numerous (eg, interfaces, particles, solid solution, clusters, and dislocations) and the relevance of the tool(s) to be used is tightly linked to the question to be answered. First, essential definitions are introduced in the form of questions: What is randomness? What is clustering? What is an interface? Then methodologies are presented, which are classified as a function of the feature they can reveal: solid solution analysis, clusters, precipitates/particles, and interfaces. The way available algorithmic methods are built is presented. Attention is focused on the reliability of the extracted information, based on either a mathematical description of the data set or on the application of data mining tools to simulated data sets. In the second part of this chapter, ways to show evidence of (and to eventually correct for) evaporation or compositional artifacts are introduced.