Total weed control within a crop is both difficult and expensive to achieve, so that some weeds will often remain to set seed. The seed production resulting from these weeds will ultimately affect the sustainability of the weed control strategy. If too much is allowed to return each season there could be a gradual, but significant, increase in the potential weed flora over a number of seasons. Field trials were carried out in 2000 and 2001 to quantify the potential magnitude of this weed seed return from Chenopodium album L., grown at two planting densities either in pure stands or in competition with one of two crops (cabbage or onion). Crop and weed weights and weed seed production were notably greater in 2001. Both dry weight and seed production of C. album were suppressed by increasing planting density or by the presence of crop, with cabbage having a more suppressive effect. Despite the plasticity in seed production, a linear relationship was demonstrated between log weed seed production and log weed biomass that was robust over a range of competitive situations with onion and cabbage, at different planting densities and in growing seasons. The study also demonstrated that the relationship could be combined with an existing simple competition model to allow the consequences of incomplete weed control to be assessed in terms of potential weed seed return.