The abundance of wolf spiders (Lycosidae) was measured across woodland–pasture boundaries in the wheat-belt of New South Wales, Australia, to determine the nature and magnitude of any edge effect. Spiders were collected by spotlighting along sample plots in woodlands located at distances of 5, 20, 35 and 200 m from the edge, and along sample plots in paddocks located at distances of 5 and 20 m from the edge. The wolf spider assemblage changed significantly across the edge, but the difference could be accounted for only by a change between the woodland and the paddock and not by any changes within the woodland at different distances from the edge. Ground cover (wolf spider microhabitat) changed significantly between the paddock and the woodland, but there were no consistent differences in microhabitat with distance from edge within either paddocks or woodlands. There was a significant correlation between an ordination of sites based on spider species abundance and an ordination based on microhabitat variables, suggesting that the wolf spider assemblage was responding to differences in microhabitat. Fine-scale selection of microhabitat by most wolf spider species was non-random, with most species preferring locations with grass cover, rather than more open locations. The present study indicates that wolf spiders are mostly unaffected by edge conditions at the woodland–paddock boundary. Accordingly, small and/or linear remnants with high edge-to-area ratios may constitute suitable faunal habitat for wolf spiders and perhaps other terrestrial arthropod species, despite the fact that this configuration is unsuitable for many vertebrate species.