The present work provides an overview of the different methods used in molecular diversity analysis. Issues like identifying voids in proprietary databases, reducing the number of redundancies present in databases, or designing focused libraries by grouping compounds similar to a template with the aim to fine tune its properties, are potent diversity analysis tools that may be used to optimize molecules based on their properties and specifically, to speed up the process of lead discovery and optimization. The present work describes first methods that are used to describe molecular systems. This is followed by a section devoted to describe different measures of similarity between molecules, to finish with a description of different methods used to select subsets molecules according to the constraints imposed. The final section deals with the validation of these methods, based on different studies available in the literature.