The aim of this study was to determine how the addition of different grasses to lucerne will affect on weed appearances and forage yield. Field experiment was set by randomized complete block design in October 2014 in Tenja (Republic of Croatia). The experiment was conduct in three repetitions on plot surface of 6 m2. Because of the late sowing date, experimental plots were covered with agrilic foil „Lutrasil“ and by the end of February the foil was removed with great overwintering of the planted plants in experiment. Floristic observations were made on April 05, 2015. The mowing was done in 2 terms: May 08, 2015 and June 16, 2015. In both mowing terms, lucerne was in the stage of flowering, while grasses were in the budding stage on first mowing term and in vegetative stage on second mowing term. By floristic analysis presence of 13 weed species were determined in experiment and dominant ones were annual broad leaf weeds. The smallest number of weeds per m2 had lucerne mixtures with large number of grasses (5th and 6th experimental treatment) – 43 and 42 weeds/m2, lucerne monoculture had 98 weeds/m2 and the largest number of weeds/m2 had esparcet, 166 of them. Highest yielding experimental treatment regarding to green mass (GM) was treatment number 5 (which included: Lucerne, Red fescue, Tall fescue, Orchard grass and Perennial Ryegrass) with yield of 50,47 t/ha of GM in 2 mowing terms, lucerne monoculture had 41,5 t/ha of green mass and smallest yield of green mass had the esparcet monoculture (23,19 t/ha GM). In terms of dry matter (DM), highest yielding was lucerne monoculture which had yield of 10,13 t/ha DM with expected annual yield of 18,41 t DM/ha. The remaining experimental treatments had dry matter yield between 14,04 and 16,85 t/ha how much had the mixture of lucerne and tall fescue, while the smallest yield of dry matter, same as the green mass, had the esparcet monoculture with 4,66 t DM/ha in 2 mowing.