The relationship between sequence, structure, and function is examined by comparing nineteen cyclic nucleotide monophosphate binding domains of known structure from six different functional families. Comparisons are made by structure and sequence alignment and through the generation of 3610 homology models. This analysis suggests there are only weak relationships between functional families, sequence, and/or structure. However, we have identified that for cyclic nucleotide monophosphate binding domains privileged template structures occur for homology modeling. The existence of privileged template structures, capable of creating accurate modeling for a broad family of proteins, may lead to improved homology modeling protocols.