Until recently, phytopathogenic bacteria have not been considered potential biological weed control candidates because they lack the ability to penetrate intact plants. This deficiency can be overcome by providing entry wounds or using surfactants. Spray application of Pseudomonas syringae pv. tagetis (5 × 108cells/ml) in aqueous buffer with a surfactant produced severe disease in Canada thistle, common ragweed, Jerusalem artichoke, sunflower, and certain other members of the Compositae under field conditions. Spray application of the bacterium without surfactant was ineffective on all reported hosts. Xanthomonas campestris pv. poannua controlled annual bluegrass in bermudagrass golf greens when applied by spray during mowing. The bacterium entered through mowing injuries, causing lethal, systemic wilt. Application of the bacterium to annual bluegrass in the absence of fresh mowing injuries failed to produce symptoms. Under field conditions, this previously unknown pathovar's host range was limited to a single subspecies of annual bluegrass, but inundative application to freshly mowed turf resulted in infection of diverse annual bluegrass biotypes. In field trials, six monthly applications resulted in greater than 70% control. The preceding examples are among the first attempts to use foliar phytopathogenic bacteria for biological weed control. Efficacy of these bacterial bioherbicides and of future biocontrol strategies employing bacteria is dependent on facilitated host penetration.