Hits is a free database devoted to protein domains. It is also a collection of tools for the investigation of the relationships between protein sequences and motifs described on them. These motifs are defined by an heterogeneous collection of predictors, which currently includes regular expressions, generalized profiles and hidden Markov models.