Grasslands provide valuable resources and refuges that support arthropod biodiversity in everchanging agricultural landscapes. Potential exists to manipulate and manage grasslands to enhance the efficiency of natural biocontrol services provided to surrounding croplands, whilst discouraging the build-up of agricultural pests. In order to successfully achieve this, a deeper understanding of the ecological function of grassland habitats is needed. Grasslands are diverse habitats that differ significantly in their composition, function and management, despite consistently being bundled together in research. Exploring these differences and how they influence arthropod communities, and thus landscape pest suppression, remains a significant gap in biocontrol research. Current knowledge of the arthropod pests and natural enemies that utilise different grassland habitats is scarce, and even less is understood about the potential for different grasslands to be biological control service providers, making progress in this area challenging. Three distinct grassland types were distinguished in the mixed farming regions of southeast Queensland, Australia. Native grasslands comprised a mix of predominantly native grass species, Mixed Improved grasslands were a sown mix of both introduced and native grass species, and Pure grasslands comprised a single dominant grass species. To compare the above-ground arthropod communities that inhabit these grasslands, field sampling was carried out over a two-year period. Arthropods were collected from grassy fields by sweep netting, while vegetation assessments monitored grassland composition, floristic diversity, quality and management practices. Only 6% of all arthropods collected were agriculturally important pests, while 25% were beneficial predators and parasitoids. Arthropods were almost 50% more abundant in Native grasslands than in the other types, although Pure grasslands supported the largest communities of natural enemies. Pest and other herbivorous insects favoured grasslands with higher proportions of weeds and other undesirable vegetation (i.e. Native and Mixed), and were less abundant in grasslands with high proportions of desirable introduced grass species (i.e. Pure). Natural enemies used all grasslands as a refuge habitat when adjacent crops were absent, while pest and other herbivorous arthropods did not. This study highlights the importance of grasslands as valuable arthropod refuges, particularly for beneficial natural enemies, and draws attention to the differences between different grassland types that can influence these communities. The observation that grasslands harbour significant communities of predatory and parasitic arthropods raises the question of whether these natural enemies are providing biological control benefits to the greater landscape. An experiment using sentinel prey was conducted to test the hypothesis that grasslands do provide pest suppressive services to surrounding crops, although iii services provided by different grassland types will differ...