Ground beetles are well known as beneficial organisms in agroecosystems, contributing to the predation of a wide range of animal pests and weed seeds. Tillage has generally been shown to have a negative effect on ground beetles, but it is not known whether this is because of direct mortality or the result of indirect losses resulting from dispersal caused by habitat deterioration. In 2005, field experiments measured direct, tillage-induced mortality, of four carabid weed seed predators, Harpalus rufipes DeGeer, Agonum muelleri Herbst, Anisodactylus merula Germar, and Amara cupreolata Putzeys, and one arthropod predator, Pterostichus melanarius Illiger, common to agroecosystems in the northeastern United States. Three tillage treatments (moldboard plow, chisel plow, and rotary tillage) were compared with undisturbed controls at two sites (Stillwater and Presque Isle) and at two dates (July and August) in Maine. Carabid activity density after disturbance was measured using fenced pitfall traps installed immediately after tillage to remove any effects of dispersal. Rotary tillage and moldboard plowing reduced weed seed predator activity density 52 and 54%, respectively. Carabid activity density after chisel plowing was similar to the undisturbed control. This trend was true for each of the weed seed predator species studied. However, activity density of the arthropod predator P. melanarius was reduced by all tillage types, indicating a greater sensitivity to tillage than the four weed seed predator species. These results confirm the need to consider both direct and indirect effects of management in studies of invertebrate seed predators.