Earthworm species contribute to soil ecosystem functions in varying ways. Important soil functions like structural maintenance and nutrient cycling are affected by earthworms, thus it is essential to understand how arable farm management influences earthworm species. One aim of arable field margin strips and non-inversion tillage is to enhance agrobiodiversity, however their influence on earthworm species assemblages remains unclear. In particular, on-farm studies conducted over multiple years that capture variability across the landscape are rare. The current study monitored earthworm species assemblages on 4 farms in Hoeksche Waard, The Netherlands, from 2010 to 2012. It was hypothesised that arable field margin strips (FM) and non-inversion tillage (NIT; a reduced tillage system that loosens subsoil at 30-35 cm depth) would have higher earthworm species abundances (epigeics and anecics in particular), soil organic matter, and soil moisture than adjacent mouldboard ploughing (MP) fields, and that earthworm numbers would decrease with distance away from FM into arable fields (MP only). FM contained a mean total earthworm abundance of 284 m-2 and biomass of 84 g m-2 whereas adjacent MP arable fields had only 164 earthworms m-2 and 31 g m-2. Aporrectodea rosea, Lumbricus rubellus, Lumbricus terrestris, and Lumbricus castaneus were significantly more abundant in FM than adjacent arable soil under MP. However, no decreasing trend with distance from FM was observed in earthworm species abundances. A tillage experiment initiated on the farms with FM showed that relative to MP, NIT significantly increased mean total earthworm abundance by 34% to 275 m-2 and mean total earthworm biomass by 15% to 51 g m-2 overall sampling dates and farms. L. rubellus, A. rosea, and L. terrestris were significantly more abundant overall in NIT than MP. FM and NIT positively affected earthworm species richness and abundances and it is noteworthy that these effects could be observed despite variation in environmental conditions and soil properties between samplings, farms, and crops. Higher top-soil organic matter and less physical disturbance in FM and NIT likely contributed to higher earthworm species richness and abundances. The anecic species L. terrestris (linked to water infiltration and organic matter incorporation) was more abundant in FM, but densities remained very low in arable soil, irrespective of tillage system.