We carried out a landscape experiment in eight 5x5 km quadrates in the Mezőföld region of Hungary. We studied how the proximity of grassland habitat patches, and at a larger scale the amount of grassland and other non-crop habitats in the landscape affect natural enemy in this case spider populations in arable fields. We found a strong effect of proximity: cereal fields had nearly twice the abundance and species number of spiders if they had a neighbouring grassland habitat patch, as op-posed to remote fields in pure agricultural settings. On the other hand, when we compared remote fields across landscape quadrates, we could not show any significant correlation between the ratio of various non-crop habitats in the quadrates and spider community parameters. The study proves the usefulness of the inclusion of natural habitat patches in a landscape, and underlines the importance that they should be interspersed with fields.