Weed management is a major constraint in organic cropping systems. In 2004, the Cornell Organic Vegetable Cropping Systems Experiment was established in central New York state using a split-plot randomized complete block design with two crop rotation entry points (split-plot factor). Four organic vegetable cropping systems that varied in cropping intensity and tillage (main plot factor) were compared: (1) intensive, (2) intermediate, (3) bio-extensive, and (4) ridge tillage. The basic crop rotation was cabbage, lettuce, potato, and winter squash, with additional short-season crops in the intensive system and with cover crops and fallow substituted for cabbage and potato in the bio-extensive system. In 2014, two uniformity trials were conducted in which oat and then a mixture of sorghum-sudangrass plus Japanese millet were grown uniformly over the entire experiment. Prior to sowing oat, soil samples were collected from each plot and an emergence bioassay was conducted to assess the soil weed seedbank. Crop biomass, weed density, and weed biomass were sampled in the uniformity crops. Soil weed seedbank density was three to four times greater in the intensive, intermediate, and ridge-tillage systems than in the bio-extensive system. The bio-extensive system also had lower weed density and weed biomass in the oat uniformity trial compared with the other three systems. Oat biomass did not differ between the cropping systems. Weed density and biomass in oat were also affected by the crop rotation entry point. Cropping system legacy effects on weed abundance and community composition were greater in the oat than in the sorghum-sudangrass plus Japanese millet uniformity trial. Our results illustrate the effects of different organic vegetable production practices on weed community structure and highlight the value of tilled fallow periods, cover crops, and prevention of weed seed rain for reducing weed populations.