Growth chamber studies and field trials were conducted to evaluate the toxicity of glyphosate [N-(phosphonomethyl)glycine] and certain other herbicides to wild garlic (Allium vineale L.). Toxicity to wild garlic increased with increasing rates up to 6 kg/ha. Glyphosate was more toxic to wild garlic when applied in spray volumes of 125 to 500 L/ha than in greater volumes. Addition of surfactant did not increase toxicity or absorption rates of 3 kg/ha of glyphosate, but did increase the advantage of 2 kg/ha of glyphosate over 2,4-D [(2,4-dichlorophenoxy)acetic acid]. When glyphosate was sprayed on wild garlic plants at 3 kg/ha, then washed off after 2, 4, 8, 16, and 32 h, absorption of a biologically significant amount of glyphosate required 8 h, and toxicity increased through the 32-h exposure. Injury, which increased as mowing was delayed up to 12 days after application, indicated that significant quantities of glyphosate were translocated between 3 to 12 days after treatment in a field test. Glyphosate; 2,4-D; 2,4-D + mecoprop {2-[(4-chloro-o-tolyl)oxy] propionic acid} + dicamba (3,6-dichloro-o-anisic acid); and VEL-4207 [N-phenyldiethanolamine-bis-(2-methoxy-3,6-dichlorobenzoate)], applied in a sequence of either spring-fall-spring or fall-spring, were compared for wild garlic control. All herbicides tested controlled wild garlic, but glyphosate was more effective than 2,4-D after spring-fall-spring applications. All spring-fall-spring treatments except 2,4-D + mecoprop + dicamba significantly reduced the number and weight of hardshell bulbs present 8 weeks after the last treatment. The initial treatment in the spring-fall-spring sequence was necessary to reduce hardshell bulb populations the second spring. One year after treatment wild garlic reinfested all plots.