The within-field spatial distribution of beneficial arthropods was assessed using two-dimensional grids of pitfall traps and suction sampling across two winter wheat fields of 4 and 16 ha, before and after an application of dimethoate. An unsprayed 6 m wide buffer zone was left around half the edge of the larger field. Arthropod numbers fluctuated to varying extents prior to spraying. Two species of Carabidae (Pterostichus madidus and P. melanarius), Linyphiidae, Lycosidae and Aphidius spp. (Braconidae) all showed their greatest reduction after spraying. For five carabid taxa, Tachyporus spp. (Staphylinidae) and Collembola the decline in numbers following spraying was no greater than any reduction found during the pre-spraying period. Within field spatial distributions of three arthropod groups were analysed using SADIE. P. madidus, present in patches across the centre of both fields prior to spraying, was removed by dimethoate and by 34 days after spraying had recovered most at the field edges. Linyphiidae were evenly distributed across both fields prior to spraying. Their numbers were reduced considerably by dimethoate and they did not recover to pre-spray levels. However, where recovery occurred this was across the centre of both fields indicating their potential to reinvade whole fields. Aphidius species were also evenly distributed across both fields prior to spraying, but did not recover after spraying. Some, but not all arthropods survived within the unsprayed buffer zone and there was some indication that reinvasion of the mid-field was more extensive where this was present. The importance of field margins with respect to insecticide treatments is discussed.