ABSTRACT Information about the impact of herbicides in the soil based on the growth of bioindicator species is extremely useful in developing crop management strategies. Therefore, this study aims to evaluate the leaching potential of the herbicide S-metolachlor under different natural precipitations in medium-textured Oxisol using bioindicator plants. A completely randomized experimental design was adopted, with four replicates and treatments arranged in a 3 × 8 factorial scheme [three indexes of precipitation occurred in the environment before the collection of the samples (50, 91, and 131 mm) and eight depths in the soil profile (0-0.03; 0.03-0.06; 0.06-0.09; 0.09-0.12; 0.12-0.15; 0.15-0.20; 0.20-0.25; 0.25-0.30 m)]. PVC columns were used, maintaining the original soil integrity during sampling after accumulating the stipulated natural precipitation. Longitudinal sections separated the columns to sow the bioindicator species (cucumber, lettuce, Alexander grass, and sorghum). The phytotoxicity symptoms of bioindicator plants were evaluated, adopting a phytotoxicity visual scale between 0 and 100%, at 5, 7, 9, and 11 days after seeding. The responses of the bioindicator species to the residual effect of the herbicide S-metolachlor were variable and depended on the rainfall level. Generally, in a medium-textured Oxisol, the higher values of concentration of S-metolachlor occurs in depths ranging between 0 and 0.06 m. The maximum leaching depth detected was 0.12-0.15 m with 131 mm of precipitation. Cucumber was the most sensitive species to the presence of S-metolachlor in an Oxisol of medium-texture since it presents symptoms of phytotoxicity at higher depths.