CellPhoneDB v2.0 is a database that infers cell-cell communication from the combined expression of multi-subunit receptor-ligand complexes. It integrates a statistical framework to predict enriched cellular interactions between two cell types using single-cell transcriptomics data. The database includes detailed information on ligands, receptors, and their interactions, with a focus on multi-subunit complexes. It allows users to introduce new interacting molecules and provides tools for analyzing large datasets efficiently. CellPhoneDB v2.0 is publicly available at https://github.com/Teichlab/cellphonedb and as a web interface at http://www.cellphonedb.org/. The database enables the identification of biologically relevant interactions from scRNA-seq data, using statistical methods to determine cell-type specificity. It supports visualization tools such as dot plots and heatmaps. The database is built on an SQLite relational database and includes a Python package for analysis. It also allows users to submit their own curated datasets. The statistical inference method uses permutations to calculate p-values for receptor-ligand interactions, identifying significant interactions between cell types. The database also includes subsampling capabilities to accelerate analysis of large datasets. CellPhoneDB has been applied to study maternal-fetal communication, cell-cell interactions in asthma, and tumor microenvironments. It provides a framework for analyzing cell-cell communication in various biological contexts. The database is supported by a range of scientific collaborations and is designed for both researchers and users to explore and analyze cell-cell communication networks.CellPhoneDB v2.0 is a database that infers cell-cell communication from the combined expression of multi-subunit receptor-ligand complexes. It integrates a statistical framework to predict enriched cellular interactions between two cell types using single-cell transcriptomics data. The database includes detailed information on ligands, receptors, and their interactions, with a focus on multi-subunit complexes. It allows users to introduce new interacting molecules and provides tools for analyzing large datasets efficiently. CellPhoneDB v2.0 is publicly available at https://github.com/Teichlab/cellphonedb and as a web interface at http://www.cellphonedb.org/. The database enables the identification of biologically relevant interactions from scRNA-seq data, using statistical methods to determine cell-type specificity. It supports visualization tools such as dot plots and heatmaps. The database is built on an SQLite relational database and includes a Python package for analysis. It also allows users to submit their own curated datasets. The statistical inference method uses permutations to calculate p-values for receptor-ligand interactions, identifying significant interactions between cell types. The database also includes subsampling capabilities to accelerate analysis of large datasets. CellPhoneDB has been applied to study maternal-fetal communication, cell-cell interactions in asthma, and tumor microenvironments. It provides a framework for analyzing cell-cell communication in various biological contexts. The database is supported by a range of scientific collaborations and is designed for both researchers and users to explore and analyze cell-cell communication networks.