The conformational agenda harnessed by different glycosidases along the reaction pathway has been mapped by X-ray crystallography. The transition state(s) formed during the enzymic hydrolysis of glycosides features strong oxocarbenium-ion-like character involving delocalization across the C-1-O-5 bond. This demands planarity of C-5, O-5, C-1 and C-2 at or near the transition state. It is widely, but incorrectly, assumed that the transition state must be (4)H(3) (half-chair). The transition-state geometry is equally well supported, for pyranosides, by both the (4)H(3) and (3)H(4) half-chair and (2,5)B and B(2,5) boat conformations. A number of retaining beta-glycosidases acting on gluco -configured substrates have been trapped in Michaelis and covalent intermediate complexes in (1)S(3) (skew-boat) and (4)C(1) (chair) conformations, respectively, pointing to a (4)H(3)-conformed transition state. Such a (4)H(3) conformation is consistent with the tight binding of (4)E- (envelope) and (4)H(3)-conformed transition-state mimics to these enzymes and with the solution structures of compounds bearing an sp (2) hybridized anomeric centre. Recent work reveals a (1)S(5) Michaelis complex for beta-mannanases which, together with the (0)S(2) covalent intermediate, strongly implicates a B(2,5) transition state for beta-mannanases, again consistent with the solution structures of manno -configured compounds bearing an sp (2) anomeric centre. Other enzymes may use different strategies. Xylanases in family GH-11 reveal a covalent intermediate structure in a (2,5)B conformation which would also suggest a similarly shaped transition state, while (2)S(0)-conformed substrate mimics spanning the active centre of inverting cellulases from family GH-6 may also be indicative of a (2,5)B transition-state conformation. Work in other laboratories on both retaining and inverting alpha-mannosidases also suggests non-(4)H(3) transition states for these medically important enzymes. Three-dimensional structures of enzyme complexes should now be able to drive the design of transition-state mimics that are specific for given enzymes, as opposed to being generic or merely fortuitous.