Dedicated to Albert Eschenmoser with best wishes on the occasion of his 75th birthday Chiral, enantiomerically pure Pd-catalysts were used to control the regioselectivity of nucleophilic attack in allylic substitutions with optically active 1,3-disubstituted allyl acetates (Schemes 4 ± 6). In contrast to reactions with achiral catalysts, where the regioselectivity is determined by the steric and electronic effects of the allylic substituents, chiral catalysts allow selective preparation of either one of the two regioisomeric products, depending on which enantiomer of the catalyst is employed. It is not necessary to start from an enantiomerically pure substrate, because the major and minor enantiomers are converted to different regioisomers (not to enantiomeric products; see Scheme 3), resulting in products of very high ee, even when the starting material is only of moderate enantiomer purity.Introduction. ± Palladium-catalyzed allylic substitutions belong to the standard repertoire of modern organic synthesis. Mild conditions, compatibility with many functional groups, and the option to control the reactivity and selectivity of the catalyst by complexing the Pd-atom with a specific ligand are attractive features, distinguishing these reactions from ordinary nucleophilic substitutions [1]. During the last few years, highly effective chiral ligands have been developed that can induce very high enantioselectivities in Pd-catalyzed allylic substitutions [2]. While high levels of enantiocontrol have been achieved in many cases, regiocontrol often remains a problem.With substrates of type 1, for example, the nucleophile can attack at either end of the allyl system and, with an achiral Pd-catalyst, the regioselectivity is determined by the steric and electronic effects of the allylic substituents R 1 and R 2 (Scheme 1). If the two substituents are similar, a ca. 1 : 1 mixture of regioisomers 2 and 3 is obtained, whereas if they are sterically or electronically distinct, moderate to high regioselectivities may be achieved. However, the regioisomer resulting from nucleophilic attack at the less reactive terminus is not accessible by this method and, therefore, it would be highly desirable to have at hand a catalyst that directs the nucleophile selectively to the desired allyl terminus, irrespective of the nature of the allylic substituents.