Field and laboratory work was carried out, from October 1982 to March 1985, to identify factors which limit the effectiveness of Demetrias atricapillus (Coleoptera:Carabidae) as a predator of cereal aphids. High overwintering densities were found in those field boundaries with a high density of tussock-forming grasses. Regular vacuum net samples of cereal fields showed that D. atricapillus began to disperse from overwintering sites in April/May, but that such dispersal was limited in extent. Field densities were highest in May and early June; accurate assessments of density required the use of an 'absolute search' method, the vacuum net being found to be only 40% efficient for D. atricapillus. Under constant laboratory conditions, rates of consumption of aphid prey were low, less than 10% of predator body weight. Estimates of predation rates in the field were obtained using data from gut dissection of field-caught beetles. The amount of solid food remains in the foregut, expressed as a Crop Volume Index, was significantly related to the biomass of food eaten. This relationship could be extended to provide estimates of the number of aphids consumed during the previous 48 hours. Such predation levels were of the order required to limit aphid population increase under certain conditions. D. atricapillus showed a diurnal bimodal activity pattern. Beetles, particularly females, readily climbed plants; at dusk over 10% of the field population were on cereal plants. Aphids were attached on plants unless prey density on the ground was high. It was concluded that poor dispersal from field boundaries in spring was the main limiting factor. Agronomic practices may also reduce the effectiveness of this species, particularly removal of boundaries or changes in their suitability for predator overwintering.