The microenvironment surrounding the active sites in polymer-supported reagents can be tailored for maximized kinetics and yields in organic reactions. Results with the Mitsunobu reaction are presented. Cross-linked copolymers of poly(vinylbenzyl chloride) are substituted with diphenylphosphine ligands at 18, 40, 67, and 100% substitution and used under Mitsunobu conditions to probe benzyl benzoate formation. It is found that the choice of groups surrounding a ligand can be as important as the choice of ligand. Decreasing the percent substitution while increasing the number of unsubstituted phenyl rings directly bonded to the polymer backbone increases the percent alcohol conversion at a 0.1 h contact time (41.7, 68.4, 83.7, and 94.5% conversion for polymers at 100, 67, 40, and 18% substitution). Polymers at 18 and 40% substitution give an equilibrium solution that is purer (97.6 and 97.0% ester) than that with a comparable soluble reagent (85.3% eater). The rapid conversion and high yield obtained as the percent substitution decreases is not due to a dilution effect : replacing the phenyl rings in the polymer at 18% substitution with carbomethoxy groups yields a polymeric reagent which allows for only 1.3% alcohol conversion at 0.1 h and a maximum product yield of 29.8%. Replacing the ester groups with the more strongly hydrogen-bonding carboxylic acid groups results in no conversion of alcohol. Thus, increasing reactant conversion with decreasing degree of substitution on a polystyrene support is a microenvironmental effect of the less polar aromtic rings superimposed on the inherent electronic effect of the CH(2)PPh(2) ligand. It is proposed that decreasing the polarity of the microenvironment surrounding the active sites increases the reactivity of the benzoate/phosphonium ion pair and lowers the energy of the S(N)2 transition state (due to the accompanying charge dispersal, as described by the Hughes-Ingold theory), resulting in an increase in the rate of product formation.