Landscape ecology has highlighted the role of the surrounding landscape in structuring arthropod communities at a given place. Previous studies have confirmed that non-crop habitats in agricultural landscapes increase the abundance and diversity of natural enemies in crop fields, which can generate a positive impact on pest control. However, landscape structure may differentially affect generalist predator communities within a year. Our knowledge about how such intra-annual variation affects the biological control of pests is still scarce. In this study, we surveyed predator carabid and spider assemblages and estimated biological control potential using experimentally added aphids in 21 winter wheat fields from mid-March until harvest. We investigated whether landscape structure differentially affected predator activity density, species richness and species composition across months, as well as how the relationship between biological control potential and landscape structure changed over time. We also investigated whether different time steps (15-day sampling period vs. whole sampling season) influenced the detection and significance of landscape effects, Spider assemblages benefit from tine-grained landscapes in spring, but this effect disappeared later in summer, For carabids, species richness was the only biodiversity measure related to grain size, and the relationship was strongest when considering the whole sampling season, No relationship was found between landscape structure and biological control potential. Our study showed that the influence of landscape structure can be inconsistent across taxa, biodiversity measures and time. The results demonstrate the importance of time steps and seasonality in the relationship between landscape structure and biodiversity.