The paper introduces SplitsTree, an interactive program for analyzing and visualizing evolutionary data. It addresses the challenge of real evolutionary data containing multiple and sometimes conflicting phylogenetic signals, which can make it difficult to support a unique tree. The method of split decomposition, developed by Bandelt and Dress, is implemented in SplitsTree to transform evolutionary data into a sum of 'weakly compatible splits' and represent them as a split graph. For ideal data, this results in a tree, while less ideal data produce a tree-like network indicating different and conflicting phylogenies. SplitsTree supports various distance transformations, the computation of parsimony splits, spectral analysis, and bootstrapping. It is available in two versions: an interactive Macintosh version (shareware) and a command-line Unix version (public domain). The program includes features for preprocessing data, such as excluding certain taxa or sites, and offers a range of methods to compute split systems, including split decomposition, parsimony splits, and spectral analysis. The paper also provides a detailed description of the program's menus, windows, and file format, along with examples and references to biological applications of the split decomposition method. The authors acknowledge the contributions of various researchers and funding sources that supported the development of SplitsTree.The paper introduces SplitsTree, an interactive program for analyzing and visualizing evolutionary data. It addresses the challenge of real evolutionary data containing multiple and sometimes conflicting phylogenetic signals, which can make it difficult to support a unique tree. The method of split decomposition, developed by Bandelt and Dress, is implemented in SplitsTree to transform evolutionary data into a sum of 'weakly compatible splits' and represent them as a split graph. For ideal data, this results in a tree, while less ideal data produce a tree-like network indicating different and conflicting phylogenies. SplitsTree supports various distance transformations, the computation of parsimony splits, spectral analysis, and bootstrapping. It is available in two versions: an interactive Macintosh version (shareware) and a command-line Unix version (public domain). The program includes features for preprocessing data, such as excluding certain taxa or sites, and offers a range of methods to compute split systems, including split decomposition, parsimony splits, and spectral analysis. The paper also provides a detailed description of the program's menus, windows, and file format, along with examples and references to biological applications of the split decomposition method. The authors acknowledge the contributions of various researchers and funding sources that supported the development of SplitsTree.