Evaluation was carried out for the interrelation between maize and Urochloa brizantha cv. Marandu and Urochloa ruziziensis, with and without application of glyphosate underdoses, and the effects of this management on weeds. For each forage species, an experiment was conducted in randomized block design, with subdivided plots. The main factor consisted of four cropping systems: intercropping of maize and U. brizantha, treated with 0, 50 and 100 g a.e. ha-1 of glyphosate, and maize monoculture; and for the intercropping of maize and U. ruziziensis, the doses of glyphosate applied were 25 and 50 g a.e. ha-1. The secondary factor consisted of five evaluation periods, with dry mass of maize and forage plants being measured on the day when the herbicide was applied, 15 days after the application, at full tasseling of maize, in hard farinaceous grain stage, and at maize harvest. Weed density and dry mass were also evaluated. The intercropping system reduced the density and dry mass production of the weed community, without affecting the maize yield components. Glyphosate suppressed the initial growth of U. brizantha and U. ruziziensis intercropped with maize, at the dose of 100 and 50 g a.e. ha-1, respectively, without compromising straw (residues) formation.