Field experiments were conducted in 2000 to 2002 near New Deal, TX to compare weed control systems in glyphosate-tolerant cotton using mechanical cultivation, a conventional hooded sprayer, and a light-activated hooded sprayer. Data collected during the first two years indicate that the use of a light-activated weed sprayer on the Texas Southern High Plains can provide weed control similar to a conventional broadcast followed by hooded application. In 2000, weed control systems using a light-activated sprayer controlled Palmer amaranth (Amaranthus palmeri) 88 to 91%, common cocklebur (Xanthium strumarium) 87 to 91%, and silverleaf nightshade (Solanum elaeagnifolium) 86 to 96% at the end of the season. In 2001, Palmer amaranth, common cocklebur, and silverleaf nightshade were controlled 80 to 94%, 80 to 93%, and 71 to 76%, respectively, at the end of the season. In general, control achieved from glyphosate postemergence-topical followed by the light-activated sprayer was similar to the conventional glyphosate applications (postemergence-topical followed by postemergence-directed), but greater than control received from two applications using the light-activated weed sprayer only. Herbicide savings ranged from 56 to 85% in 2000 and 63 to 84% in 2001. Weed maps at application were created in 2001 and 2002 to monitor changes in weedy patches over time. Hyperspectral images were created to examine differences in wavelength reflectance between cotton (Gossypium hirsutum), peanut (Arachis hypogaea), and a variety of broadleaf weeds. In addition, differences in wavelength reflectance within a weed species at different growth stages and health conditions is currently being examined.