CLARE is a computational method designed to discover novel tissue-specific regulatory elements. Starting with a set of regulatory elements known to be active in a particular tissue/process, it learns the sequence code of the input set and builds a predictive model from features specific to those elements. The resulting model can then be applied to user-supplied genomic regions to identify novel candidate regulatory elements. CLARE's model also provides a detailed analysis of transcription factors most likely to bind the elements, making it an invaluable tool for understanding mechanisms of tissue-specific gene-regulation.