PrePPI is a database of predicted and experimentally determined protein-protein interactions (PPIs) for yeast and human. Predicted interactions are assigned a likelihood using a Bayesian framework that combines structural, functional, evolutionary, and expression information. The database contains ~2 million predictions including 31,402 for yeast and 317,813 for human that are considered high confidence based on our analysis. Experimentally determined interactions are compiled from a set of public databases (e.g., DIP, IntAct, HPRD, etc.).