This article reports experiments that explore the critical factors for gene flow in oilseed rape (Brassica napus) via the soil seedbank under Chinese farming systems. Research was carried out on 16 farmerÕs fields to estimate harvest seed losses after different planting methods and harvest timings, during the four components of the harvest operations. An additional field experiment was set up to estimate the potential germination proportion of seeds in flood-irrigated and unirrigated fields, as a function of seed burial depth over time. Seed losses were largest during the sun-drying of harvested plants in the field, accounting for about three quarters of the total seed losses. The number of intact seeds declined quickly after they entered the soil, particularly in un-irrigated fields, and no volunteers germinated in the irrigated field after 3 months. Burial depth did not affect the potential for germination and emergence of volunteers in the flood-irrigated field. However, the potential germination declined with increasing burial depth in the un-irrigated field. No viable seeds remained in the soil after 16 months. Seed losses during harvest could be limited to 0.7-1.1% by weight of total seed production by placing a plastic membrane under the cut plants during drying and no rape seeds could germinate after 3 months in waterlogged condition. Thus, the risk of gene flow via the soil seedbank is probably small in Chinese oilseed rape farming systems.