Summary Habitat restoration is commonly conducted in agricultural landscapes with the aim of restoring biodiversity. Some species, however, might not be able to migrate to restored habitats, and vital habitat elements, such as logs, may be missing. We compared arthropod assemblages under logs amongst different land-use types: pastures, revegetation and woodland remnants, in south-eastern Australia. We also supplemented habitats with logs, placed out in different seasons and for different periods of time to determine how season and exposure time affect arthropod composition. We compared assemblages under logs in revegetation adjacent to and isolated from woodland remnants to test the role of habitat connectivity. Arthropod assemblages differed significantly between land-use types, with pastures most different to remnants. These differences corresponded with differences in log microhabitat. Time was an important determinant of community composition, but habitat (remnant vs revegetated) and revegetation connectivity (adjacent vs isolated patch) were not. Time affected assemblage composition in two distinct ways: first, time of year (November vs January), and second, exposure time of logs (1 vs 3 months) affected composition. Exposure time effects may indicate dispersal limitation, but the proportion of wingless species did not depend on exposure time or connectivity. We conclude that the log fauna in this landscape responds to microenvironments and seasonal change but is not strongly dispersal limited, allowing it to respond rapidly to habitat restoration. The pre-agricultural landscape likely shared many features with the modified landscape, such that many species possess traits and behaviours that allow them to move through and persist in the matrix.