Barnyardgrass was grown at densities of 0, 0.25, 0.5, 1, 2, 5, and more than 50 plants m−1 of tomato crop row in either a regular, random, or clumped pattern. Tomato was established at 0, 5, 10, or 20 plants m−1 of crop row in a regular pattern. Crop density and weed density or spatial arrangement had little effect on phenological development of barnyardgrass. In the absence of tomato, barnyardgrass was estimated to produce over 400,000 seeds plant−1 when not subjected to intraspecific competition (0.25 plants m−1 density), decreasing to about 10,000 seeds plant−1 when weed density exceeded 50 plants m−1 of row. Differences in seed production between plants in the regular and random spatial arrangements were minor, but the clumped distribution resulted in 30 to 50% reduction in seed production at weed densities between 1 and 5 plants m−1 of row. Tomato reduced barnyardgrass seed production. The magnitude of the reduction depended on both tomato density and barnyardgrass density. In the absence of tomato, barnyardgrass produced over 200,000 seeds m−2 in 1993 and over 500,000 seeds m−2 in 1994 at 5 plants m−1 of row. Production was almost 700,000 seeds m−2 when the weed density exceeded 50 plants m−1 of row. Barnyardgrass seed production at the single-season economic threshold density in tomato was sufficient to maintain the seedbank at a level that would mandate high levels of weed control in subsequent crops. Because of the high fecundity of barnyardgrass, our experiments suggest that stopping seed production is the best long-term management strategy for the weed. Spatial arrangement of the weed, at the scale used in these studies, would not be a factor in establishing long-term management guidelines based on weed population biology.