ABSTRACT The biological control, thinking about the integrated management, has been inserted with other management techniques to disease control, such as soybean anthracnose. The aims of this work were to verify the action of Trichoderma and Bacillus isolates in the induction of soybean resistance mechanisms to anthracnose as a function of seed treatment. The statistical design was entirely randomised, in a 5 x 2 (agent species x sampling times) factorial scheme with five replicates. Soybean seeds were treated with Bacillus amyloliquefaciens BV03, B. subtilis BV02, Trichoderma asperellum BV10, Carbendazim + Thiram and distilled water (control). Seven days after seedling emergence, 2 μL of 1 x 10−4 Colletotrichum truncatum spores were inoculated on the cotyledons. Catalase (CAT), peroxidase (POX), phenylalanine ammonia lyase (PAL) and glyceollin (GLY) activities before and after pathogen inoculation, as well as the diameter of the anthracnose lesion on the cotyledons, were evaluated. Data were submitted to analysis of variance and, when significant, the mean values were compared by Fisher’s test (p < 0.05). The treatments did not influence the first sampling time before inoculation. Trichoderma asperellum BV10 increased POX and PAL activities up to 173%, while B. amyloliquefaciens BV03 increased POX activity. Glyceollin was not influenced by the treatments. The T. asperellum BV10 reduces the diameter of the anthracnose lesion by up to 61%. Thus, T. asperellum BV10 has the potential to control soybean anthracnose, improved the response defense against C. truncatum, when performed on seed treatment.