Phthalide derivatives represent an important class of naturally occurring lactone. Some of these compounds showed a biological activity and also have been used in the treatment of diseases. A lot of synthetic methods were developed and reported for preparation of phthalide derivatives as target molecules and starting materials for preparation of other natural products. Among metalmediated allylation reactions, indium-mediated allylation has gained popularity in view of the possibility of carrying out the reaction in water and showing different pattern of diastereoselectivity. The organic reaction in aqueous solution can save time and cost by evading a consuming protection and deprotection sequence for certain functional groups containing acidic moiety such as carboxylic acid and alcohol. Here we report an efficient allylation of hydroxyphthalides 1a and 1b using various allyl bromides and indium in THF in the presence of 1 equiv. of acetic acid (Scheme 1). The allylations in aqueous THF were attempted as well for comparison. For the optimization of allylation, the ratio of indium and cinnamyl bromide in the reactions was fixed in 1 : 1.5 on the assumption of reactive species of allyl indium as a sesquidimeric species. The allylation reaction of hydroxyphthalide 1a in THF (entry 1, Table 1) was slow and gave a cinnamylated phthalide in low yield even after 24 h due to incomplete reaction. By the addition of water (entry 2), 1 equiv. of 1 N HCl (entry 3) or 0.5 equiv. of acetic acid (entry 4) to THF reaction solution as a cosolvent or additives, the allylation gave a cinnamylated phthalide in quantitative yield with increased diastereoselectivity and reduced reaction time as shown in Table 1. The reaction in the presence of acetic acid is turned out to be faster than that in aqueous solution. Acetic acid rather than 1 N HCl solution was a choice of acid additive because of fast reaction rate, easy handling and better diastereoselectivity. 1 equiv. of Acetic acid was found to be optimum amount in both reaction rate and diastereoselectivity (entry 5). The reactions of 3-hydroxyphthalide 1a with various allyl bromides in THF in the presence of 1 equiv. of acetic acid were examined (Method B) and the results are shown in Table 2. The reactions also were tried in aqueous THF (THF : H2O = 4 : 1) for a comparison (Method A). Generally, the reaction under Method B is faster than that under Method A. The allylations of hydroxyphthalide 1a with various allyl bromides under two conditions gave the corresponding products in high yields in the most cases (entries 1, 2 and 46, Table 2), whereas the propargylation of hydroxyphthalide 1a with propargyl bromide in aqueous THF did not gave any product. The reaction with propargyl bromide in THF in the presence of acetic acid gave the two products, propargyl and allenyl phthalide in 46% and 27% yields, respectively (entry 3). The allylation of phthalide 1a with crotyl bromide under the reaction conditions almost did not show any diastereoselectivity (entry 5). In the reaction of allyl bromides with substituents at γ-position, only the product resulting from γattack was produced (entries 2, 5 and 6). The reaction did not seem to proceed through the oxocarbenium ion C generated from a phthalide 1a under the our present conditions but through allylation of carboxybenzaldehyde A in the equilibrium with 3-hydroxyphthalide 1a followed by lactonization. Even if internal chelation effect of indium reagent by carboxylate has been known, it is not clear that it is operative in our phthalide system under our reaction condition. We next examined allylation of 3-hydroxy-3-methylphScheme 1 Table 1. Allylation of hydroxyphthalide 1a with cinnamyl bromide