The ChEMBL collection is the chemically curated version of the most recent release of the ChEMBL database maintained by European Bioinformatics Institute. It contains bioactivity data for more than 1 million compounds, which makes it an excellent free resource of reference ligands. These ligands can be used to build and validate virtual screening workflows. All entries have been processed by the mcule structure registration system. Molecules with structural problems and inconsistencies have been filtered out.