Most studies that have explored the effect of biological control in maintaining pest populations at acceptable levels have focused on the ability of natural enemies to reduce pest abundances. The reduction in pest populations induced by natural enemies, however, is rarely studied in association with the incidence of viruses transmitted by pests, as well as its impact on crop damage and yield. Here, we performed large-scale winter field monitoring to assess the direct and indirect effects (via arthropod natural enemies) of flower strips along cereal crop margins, on (i) cereal aphid abundance, (ii) the incidence of barley yellow dwarf viruses (BYDVs) transmitted by aphids in autumn and winter, (iii) crop damage, and (iv) crop yield. In 28 cereal fields (over 2 years), we used a paired experimental design to compare a cereal field edge directly adjacent to a flower strip with an opposite (at least 50 m apart) cereal field edge adjacent to a grassy margin. Our results highlight that winter flower strips favoured the activity of ground arthropod predators in the adjacent cereal field but not aphid parasitism rate in winter. Parasitism rate only increased with the complexity of the surrounding landscape. Our results also showed that flower strips reduced both vector abundance and virus incidence in the adjacent cereal field in winter but did not affect virus symptom levels or yields. Damages were low and yield remained high in our study, which may encourage farmers to stop insecticides during the autumn-winter. This study provides a first example of how flower strips along field edges can decrease the incidence of a vector-borne virus in winter and could be used as a strategy to promote more sustainable agriculture.