The aim of this paper is to analyze yield variation of Dactylis glomerata and Festuca pratensis varieties grown on organic and mineral soil. This paper has drawn on two field experiments set up and carried out between 2010 and 2013. The experiment was conducted in two experimental stations: one in the Research Centre for Cultivar Testing in Krzyżewo and the other in the Experimental Stations for Variety Testing in Uhinin. The experiment in Krzyżewo was set up on ploughed soil, with spring barley as a forecrop. In Uhnin the experimental plots were located on peat meadow. The experimental plots were sown with varieties of Dactylis glomerata: Niva, Tukan, Amila, Crown Royale and with varieties of Festuca pratensis: Limosa, Pasja, Anturka, Amelka. The full exploitation of Dactylis glomerata varieties was due between 2012 and 2013, whereas for Festuca pratensis it was due between 2011 and 2012. In the experimental plots with the varieties of Dactylis glomerata the grass was harvested six times a year and chemical analysis of the biomass was done taking dry matter only from five cuts. The varieties of Festuca pratensis were harvested four times. Each year in the course of the experiment fresh and dry matter of each cut were weighed. The grass species and their varieties as well as the particular mowing and kind of soil where the grass was grown have an impact on the yield. On mineral soil the yield of Dactylis glomerata was higher than Festuca pratensis. On organic soil the yield of both species was similar. During the two years of experiment the highest yield among Festuca pratensis varieties was noted for Amelka whereas among varieties of Dactylis glomerata the yield was similar and differences were not statistically significant.