Spiders play a key role in forest food webs, where they regulate decomposer populations and may act as predators of pests and disease vectors. Spider community composition is determined in part by vegetation structure. Therefore, the exclusion of large mammals, such as deer and wild boar, through wildlife fencing may affect the composition of spider communities and their prey in forest ecosystems. Web-building spiders and their prey were hand-collected in plots that had been fenced for three years, as well as adjacent unfenced plots in a mixed temperate forest in north-eastern Germany. Additionally, spiders in the leaf litter were sampled in fenced and unfenced subareas by sieving litter. Wildlife fencing did not significantly affect spider densities or community composition per microhabitat. However, fencing affected the cover of different microhabitats significantly as there was a higher density of larger trees and higher leaflitter cover in fenced plots while there was a higher percentage of bare ground and deadwood in unfenced plots. Spider communities and their prey composition differed significantly between microhabitats (deciduous trees, coniferous trees, dead wood, understory vegetation, leaf litter) independent of fencing. Thysanoptera prey was mainly caught by spiders on coniferous trees and in the understory vegetation. Heteroptera prey were captured most frequently in the understory vegetation while Hymenoptera (excl. Formicidae) prey were mostly caught on deciduous trees. Several spider species showed a preference between deciduous and coniferous trees in the mixed forest. Wildlife fencing alters the vegetation structure of mixed forests and has indirect effects on spider communities and their role in forest food webs due to alteration of microhabitat availability.