The work presented in this thesis focuses on the development of base-free 1,3-dipolar cycloaddition reactions of azomethine ylides to produce homochiral pyrrolidines. Previous work within our group has shown the suitability of using 1-benzyl-4,5-dihydroimidazole as the starting point for the generation of an azomethine ylide that undergoes 1,3-dipolar cycloadditions with a range of dipolarophiles in a highly regio- and stereoselective fashion. l-Benzyl-2-imidazoline reacts with ethyl diazoacetate in the presence of Cu(acac)₂ to form an azomethine ylide that reacts with dimethyl fumarate to yield a hexahydropyrrolo[1,2-a] imidazole. Similarly, the 2-imidazoline reacts with diazoacetate and fumaronitrile in the presence of copper(II) triflate to form analogous bicyclic systems. Other Cu(II) salts were found to be less effective catalysts, as was the Rh(II) acetate dimer. In the course of the above reactions, we isolated two diastereomeric products that proved to be 1:2 adducts between the dihydroimidazole and the electron deficient alkene. These compounds also contained the hexahydropyrrolo[ 1,2-a] imidazole core and their formation can be rationalised by (i) a Michael-type addition of the dihydroimidazole onto one molecule of the electron-deficient alkene to form a 1,3-zwitterion, (ii) a 1,2-proton shift within the zwitterion to form an azomethine ylide, and (iii) 1,3-dipolar cycloaddition of the ylide onto a second alkene molecule acting as a dipolarophile. Other electron-deficient alkenes, e.g. maleate, maleimides and fumaronitrile also undergo this reaction but ethyl cinnamate and p-nitrostyrene do not. The reaction with a mixture of two alkenes allows 1:1:1 products to be isolated. In these situations maleimides act as the Michael acceptor and the second alkene as the dipolarophile. In this way we found the order of reactivity to be maleimide > fumarate/maleate ester > fiimaronitrile. The stereochemistry of the four new chiral centres was elucidated by X-ray crystallography and n.O.e. difference spectroscopy. The formation of the major diastereoisomer can be rationalised by a transition-state model in which (i) the a-carbonyl of the ylide is 'syn' to the 2-H of the dihydroimidazole, (ii) the dipolarophile approaches in an 'endo' fashion. Formation of the minor diastereomer may involve a subsequent épimérisation at the C7a position of the fused system or 'endo' mode of attack on the alternative ylide rotamer. Incorporation of a chiral centre into the imidazoline through the use of a phenyl substituent affords complete facial selectivity for the 1,3-dipolar cycloaddition reaction. Some electron-deficient systems that are unable to participate in the 1,2-proton shift (e.g. DMAD, β-nitrostyrene) form 1:2 adducts in which the product contains a 5,6-fused system. These products can be understood as arising from two sequential Michael additions to form a 1,5-zwiterion that is able to ring close onto the imidazoline iminium ion.