The objective of this study was to figure out, how sown and native grassland develops under extensive pasturing (forage structure and quality) and how far cattle can compensate differences and lower feed qualities by diet selection. Two rotational stocking systems (0 versus 70 kg N/ha) with cattle yearlings were compared. Diet selection was the main focus of research: forage horizons (each 10 cm), selected feed quality (quality estimations based on faecal nitrogen versus forage quality by enzyme-insoluble organic matter) and animal behaviour (scan-sampling). Beside this n-alkanes were tested to be used to determine the forage components grass, herbs and legumes. The forage horizons especially of overmatured forage were clearly different in terms of contained plant parts and feed quality. Horizons rich in leaves had best feed qualities. In overmatured forage cattle preferred horizons rich in leaves – mainly found at heights from 16,5 to 36,5 cm of the standing forage. Energy concentration and organic matter digestibility of the selected diet was higher than in the mean forage offered. While finishing a paddock cattle were able to ingest high qualities pretty long. Concerning animal behaviour high significant differences were found between first and last day of pasturing: While pasturing a paddock, with decreasing forage availability grazing time increased and time laying down decreased in both treatments. Forage horizons (with differences in contained plant parts / morphological stages) showed partly high differences in n-alkane concentration and structure, while those of the forage components grass, herbs and legumes were partly similar. So with distinguishing only those three forage components on inhomogeneous, partly overmatured pastures rich in species, selection of forage components could not be determined by using n-alkanes. Regarding the cattle’s daily live weight gain, in tendency no differences between treatments could be found - deviations varied between years. Beside in 2005, live weight gain averaged clearly more than 1000 g per day. Both steers and heifers reached with U and R good carcass classifications (EUROP-classification). Almost 80 % showed the optimum fat cover (class 3).