Approximately 50% of the genetically modified herbicide-resistant corn hectares in the United States are treated only with POST-applied herbicides for weed management. Although a high degree of efficacy can be obtained with POST-applied herbicides, delayed timing of application may result in substantial corn yield loss. Our goal was to characterize on-farm corn–weed communities prior to POST herbicide application and estimate potential corn-yield loss associated with early-season corn–weed competition. In 2008 and 2009, field surveys were conducted across 95 site-years in southern Wisconsin and recorded weed species, density, and height in addition to crop height, growth stage, and row spacing. WeedSOFT® was used to predict corn yield loss. Common lambsquarters, velvetleaf, dandelion, common ragweed, and Amaranthus species were the five most abundant broadleaf weed species across site-years, present in 92, 86, 59, 45, and 44% of all fields, respectively, at mean densities of 19, 3, 3, 4, and 3 plants m−2, respectively. Mean plant heights among these species were 17 cm or less. Grass and sedge species occurred in 96% of fields at a mean density of 25 plants m−2 and height of 7 cm. The mean and median of total weed density across site-years were 96 and 52 plants m−2, with heights of 14 and 13 cm, respectively. Mean predicted corn yield loss was 4.5% with a mean economic loss of $62 ha−1. However, predicted yield loss was greater than 5% on one-third of the site-years, with a maximum of 26%. These results indicate that delayed application of POST herbicides has led to corn yield loss due to early-season weed-crop competition on a substantial number of fields across southern Wisconsin, and suggest that management tactics need to be improved to protect corn yield potential fully.