Resumo. The aim of this study was to evaluated the effects of different types of weed control on the ant fauna in an initial planting of Eucalyptus grandis. Therefore, in march of 2011 was performed the implantation of culture, in an area located in the city of Santa Maria - RS. The six treatments were constituted by the total chemical control of weeds in row and interrow planting, with glyphosate (T1); total chemical control of weeds in the crop row, with glyphosate (T2); chemical control of monocotyledonous in row and interrow planting, with sethoxydim (T3); chemical control of dicotyledonous line interrow and planting, with bentazone (T4); total chemical control of weeds in a yard track parallel to the row, with glyphosate, and a meter in the central part of interrow without control (T5); and without control weeds (T6). The survey of the ant fauna was conducted in one year using three collection methods: attractive bait, trap soil and Berlese funnel, with six replicates per treatment in each collection date. In this period were collected 46,675 ants, belonging to 37 species, there being no significant difference between the total specimens collected. In the area of treatment consists of the control of weeds in the crop row only verified efficiency of 99.0% and sample S obs = 35. Among the indices of Shannon Diversity ( H ’), highlighted the value found for the area of treatment T2 ( H ’= 1.34) at the expense of the values of the areas where they were installed the T1 ( H ’= 1,25) and T5 ( H ’= 1,23). Having thus a higher co-existence of ant species between the areas with changed structure less flora, and between the areas most simplified. There were significant correlation ( r = 0.0463) between the species richness of ants collected and the number of weeds found at the end of the experiment. Thus, we conclude that the indirect effects of the herbicides affect more the local composition ant species than his richness.