It is commonly believed that sequence determines structure, which in turn determines function. However, the presence of many proteins with the same structural fold but different functions suggests that structure and function do not always correlate well. We have developed a method for accurate functional annotation, based on identification of functional signatures from structural alignments (FSSA) using the Structural Classification of Proteins (SCOP) database. SCOP is a manually curated database to classify known structures into hierarchical levels of class, fold, superfamily and family to embody structural and evolutionary relationships. We define homologues as proteins within the same SCOP superfamily, and structural analogues as proteins within the same SCOP fold but different superfamily. For each amino acid residue in a given structure, we calculate the log odds of finding similar local structure in homologues versus structural analogues. For each structure, the collection of these log odds scores comprises its FSSA signature, and this signature is used to interpret the functional importance of individual residues, or to classify a given structure into functional categories. The method is superior at function discrimination compared to several methods that directly inherit functional annotation information from homology inference, such as Smith-Waterman, PSI-BLAST, Hidden Markov Models (HMMs) and structure comparison methods, for a large number of structural fold families. Our results indicate that for proteins within multi- functional fold families, the contributions of amino acid residue to structure and to function are largely separable.