Glucansucrases are produced principally by Leuconostoc mesenteroides and oral Streptococcus species, but also by the lactic acid bacteria (Lactococci, Lactobacilli). They catalyse the synthesis of high molecular weight D-glucose polymers, named glucans, from sucrose. In the presence of efficient acceptors, they catalyse the synthesis of low molecular weight oligosaccharides. Glucosidic bond synthesis occurs without the mediation of nucleotide activated sugars and cofactors are not necessary. Glucansucrases have an industrial value because of the production of dextrans and oligosaccharides and a biological importance by their key role in the cariogenic process. They were identified more than 50 years ago. The first glucansucrase encoding gene was cloned more than 10 years ago. But the mechanism of their action remains incompletely understood. However, in order to synthesise oligosaccharides of biological interest or to develop vaccines against dental caries, elucidation of the factors determining the regiospecificity and the regioselectivity of glucansucrases is necessary. The cloning of glucansucrase encoding genes in addition to structure-function relationship studies have allowed the identification of important amino acid residues and have shown that glucansucrases are composed of two functional domains: a core region (ca. 1000 amino acids) involved in sucrose binding and splitting and a C-terminal domain (ca. 500 amino acids) composed of a series of tandem repeats involved in glucan binding. Enzymology studies have enabled different models for their action mechanism to be proposed. The use of secondary structure prediction has led to a clearer knowledge of structure-function relationships of glucansucrases. However, mainly due to the large size of these enzymes, data on the three-dimensional structure of glucansucrases (given by crystallography and modelling) remain necessary to clearly identify those features which determine function.