In the Forest District of the city of Lübeck during spring 1990 27 pairs of fenced and unfenced plots of 100 m ² were established in newly thinned beech and mixed beech-oak forests. The plots represent a range of poor forest communities on acid soils (beech and oak-beech-forests of the Luzulo-Fagion) to rich beech forests on less acid soils (ash- and maplerich variants of the Galio odorati-Fagetum). Over the years, ten of the plot pairs had to be given up. In the years 1990 to 1992, 1996 and 1997 in 17 pairs of plots relevés according to the method of BRAUN-BLANQUET were carried out. By this means we studied the impact of browsing game on the number of plant species as well as their cover percentages in the herb and shrub layers. Site quality was characterised by the indicator values according to ELLENBERG sen., in order to relate the game impact to soil parameters. The game impact on the species richness of woody species is small, but the cover of the tree regeneration in the fenced plots is significantly higher than in the unfenced ones. The totals of cover percentages over all plots of the five most important tree species differ between species. Over eight years, in the fenced plots the values of the tree species (except Quercus robur/petrea) increase. On the unfenced plots, the totals of cover of Fagus sylvatica and of Acer pseudoplatanus also increase, but significantly less so as compared to the fenced plots. For Fraxinus excelsior these totals are nearly constant over the time of investigation, the values of Carpinus betulus decrease in the unfenced plots. The frequency of the five tree species appears to stay nearly unchanged during the investigation period, only oak seems to increase slightly in the unfenced plots because of a higher number of seedlings in 1997. The average total species number and the number of woody resp. of tree species as well as their cover percentages are correlated with site quality, as defined above. Low indicator values are accompanied by a low number of species and by low cover. Species number as well as cover percentages increase with the indicator values, unless other site factors - especially irradiance - are limiting. Their reactions are more pronounced in the fenced plots as compared to the unfenced ones. - The importance of site quality for species numbers and for their cover percentages are to be considered more strongly in discussions about game impact on tree regeneration.