The concept of counteranion-directed catalysis (CDC) is applicable to reactions in which cationic intermediates are formed. [1] In principal, it can even be applied to transitionmetal-catalyzed reactions, [2] as demonstrated for gold catalysis by Toste and co-workers, [3] and for palladium catalysis by List and co-workers. [4] The latter group showed that the positively charged p-allyl/Pd II complex of the Tsuji-Trost reaction [5,6] can interact with a chiral counteranion, thus allowing an efficient direct a allylation of aldehydes by asymmetric counteranion-directed catalysis (ACDC). In the proposed mechanism, the chiral phosphate counteranion establishes hydrogen bonds with the HN moiety of the intermediate enamine as well as an ionic interaction with the charged Pd II center, thus resulting in a very organized transition state (Scheme 1).Herein, we disclose a highly diastereoselective synthesis of 2,5-disubstituted 3-hydroxy-tetrahydrofurans through the formation of a p-allyl/Pd II intermediate, in which the carbox-ylate counterion plays a prominent directing role. We propose a mechanism that is based on DFT calculations and chemical experiments, and which indicates that the stereocontrol may in part be a result of the formation of an unusual noncovalent bond between the counteranion and one of the hydrogen atoms of the cationic p-allyl/Pd complex itself. The two hydroxy groups are also involved in the mechanism, and the sum of all these noncovalent interactions leads preferentially to the highly organized chiral transition state [pR]°(see Schemes 1 and 3), a prediction that accounts well for the observed diastereoselectivity.In the course of our total synthesis of (+)-oocydin A (1; Figure 1), [7] we devised this convenient approach to 2,5disubstituted 3-hydroxy-tetrahydrofuran 4 from the readily available syn diol 3, giving 4 in a surprisingly high d.r. (trans/ cis = 96:4; Scheme 2). This high selectivity was unexpected, considering the work of Hara et al., [8] who observed a poor diastereoselectivity (trans/cis = 37:63) for the cyclization of 5 to 6 (Scheme 2). This result led us to suspect a directing effect of the b-OH group in the stereoselective cyclization of 3 to 4, and prompted us to further explore this promising reaction. The first confirmation of our hypothesis was the observation that even anti diol 7 cyclized diastereoselectively, leading to 2,5-disubstituted 3-hydroxy-tetrahydrofuran 8 [9] (anti/syn = 95:5, 97 % yield; Scheme 2). We must emphasize that in anti diols as well as in syn diols, the newly formed vinyl function is selectively installed trans to the b-OH group (which does not cyclize), while the stereogenic center that bears the g-OH group (which cyclizes) seems to have no impact on the stereoselectivity, which is counterintuitive.In order to gain some insight into the mechanism of this reaction, DFT calculations were carried out, starting with the syn diol series (Scheme 3). [10,11] The nucleophilic attack of the g-OH group at the p-allyl moiety was modeled (outer-sphere mechanism). [...