Two extracts of Sapindus saponaria were evaluated as remediators of a diesel contaminated soil, at laboratory level. The soil was contaminated with 100 ml of gasoil per kilogram and was treated with aqueous and hydroalcoholic extracts in doses of 50, 100 and 150 ml per kilogram of contaminated soil, with a control sample by natural attenuation. The design was randomized blocks with six treatments, a standard block, three replicates per treatment and a variable response (percentage of extractables with n-hexane, HEM), according to EPA 9071b. Samples were taken from each experimental unit at 15, 30 and 45 days and the HEM was determined. The statistical design was factorial ANOVA with multiple criteria of Student-Newman-Keuls and significance α = 0.05, to establish the influence of the experimental factors on the response variable. The extracts of S. saponaria had a positive effect, incre...