ABSTRACT jSymbolic is an open-source platform for extracting features from symbolic music. These features can serve as inputs to machine learning algorithms, or they can be analyzed statistically to derive musicological insights. jSymbolic implements 246 unique features, comprising 1497 different values, making it by far the most extensive symbolic feature extractor to date. These features are designed to be applicable to a diverse range of musics, and may be extracted from both symbolic music files as a whole and from windowed subsets of them. Researchers can also use jSymbolic as a platform for developing and distributing their own bespoke features, as it has an easily extensible plug-in architecture. In addition to implementing 135 new unique features, version 2.2 of jSymbolic places a special focus on functionality for avoiding biases associated with how symbolic music is encoded. In addition, new interface elements and documentation improve convenience, ease-of-use and accessibility to researchers with diverse ranges of technical expertise. jSymbolic now includes a GUI, command-line interface, API, flexible configuration file format, extensive manual and detailed tutorial. The enhanced effectiveness of jSymbolic 2.2’s features is demonstrated in two sets of experiments: 1) genre classification and 2) Renaissance composer attribution.