High-resolution characterization of protein functional sites and molecular determinants remains rooted in heavy experimental mutational analysis. JET2 Viewer is a very large knowledge base containing JET2 predictions of protein binding sites for more than 40 000 chains, occurring in Protein Data Bank (PDB) structures. These 40 000 chains represent the structural diversity of the PDB and constitute the nonredundant set of non-homologous (<40% identity) protein chains extracted from all PDB structures. JET2 is a computational method that provides very precise predictions for a wide range of protein-protein interfaces and discriminates them from small-molecule binding sites. Beyond its predictive power, the approach permits to dissect interaction surfaces and unravel their complexity. The predictions reported in JET2 Viewer are expected to help the users to foster new strategies for protein-protein interactions modulation in their systems and for interaction surface redesign.