Polysaccharide glycosyl hydrolases are a group of enzymes that hydrolyze the glycosidic bond between carbohydrates or between a carbohydrate and a noncarbohydrate moiety. Here we illustrate that traditional schemes for grouping enzymes, such as by substrate specificity or by organism of origin, are not appropriate when thinking of structure-function relationships and protein engineering. Instead, sequence comparisons and structural studies reveal that enzymes with diverse specificities and from diverse organisms can be placed into groups among which mechanisms are largely conserved and insights are likely to be transferrable. In particular, we illustrate how enzymes have been grouped using protein sequence alignment algorithms and hydrophobic cluster analysis. Unfortunately for those who seek to improve cellulase function by design, cellulases are distributed throughout glycosyl hydrolase Families 1,5,6,7,9, and 45. These cellulase families include members from widely different fold types, i.e., the TIM-barrel, beta alpha beta-barrel variant (a TIM-barrel-like structure that is imperfectly superimposable on the TIM-barrel template), beta-sandwich, and alpha-helix circular array. This diversity in cellulase fold structure must be taken into account when considering the transfer and application of design strategies between various cellulases.