In this chapter the foundations and applications of the Residual Density Analysis (RDA) are shown. The RDA is a concept for the detection and quantification of features from residual density grids. These may be from XD, MoPro, TONTO, or BayMEM. It can be used in radial function development, data processing and data reduction development, in the development of refinement strategies, as a fingerprint method for systematic errors and their imprint onto the residual density, and in day-to-day applications to the density and thermal motion models. In particular, the RDA is used as a stopping criterion in Multipole Modeling as it gives a measure for structural information (features) in the residual density distribution. When no more features are present, thermal motion and density models fit the experimental data with residuals distributed according to a Gaussian and only noise remains in the residual density. The RDA cannot give a proof for the correctness of a model, but it can disprove the expected matching with Gaussian residuals between model and data. Applications of the RDA to electron reconstructions from Multipole Models and from an application of the Maximum Entropy Method are given. Section 3.8 gives an application of the RDA as a fingerprint method.