An experiment that evaluated control of tarnished plant bug (TPB), Lygus lineolaris (Palisot de Beauvois), populations in cotton, Gossypium hirsutum L., by reduction with herbicides in numbers of broad leaf wild host plants found near fields was conducted in Sunflower and Washington Counties in the Delta of Mississippi in 1998 and 1999. Four experimental areas each approximately square and 4.8 km on a side were used in both years. Check areas received no herbicide treatment, while treated areas (one area in 1998 and two areas in 1999) received an application of Trimec® (mecoprop + 2, 4-D + dicamba) at 1.55 + 0.54 + 0.17 kg AI/ha applied with a spray system mounted on a tractor in April of both years. This herbicide kills broad leaf weeds but not grasses. In 1998, only the larger areas with good stands of wild hosts were treated with Trimec, while smaller areas with good stands of hosts were spot-treated with glyphosate (0.22 kg AI/ha) using an all terrain vehicle and hand-held wand spray system. Approximately 25% of the marginal areas with wild hosts were treated with herbicides in the treated test area in 1998. Over 90% of the marginal areas with wild hosts were treated with the tractor and Trimec in the two treated areas used in 1999. The herbicide treatment significantly reduced numbers of broad leaf hosts in sample areas in the treated areas as compared to check areas in both years. A corresponding reduction in numbers of TPB found in the sample areas in the treated areas as compared to the check areas was also found. However, in 1999 numbers of TPB found on wild hosts in the treated test areas in posttreatment samples had increased significantly as compared to their numbers found in pretreatment samples. This increase was caused by TPB utilizing annual rye grass, Lolium multiflorum Lamarck, as a host in the treated areas. In 2000, Trimec will be applied in the treated test areas in late-February and early-March to destroy the F TPB population and avoid this problem. Average numbers of TPB in cotton fields in 1999 were 3-fold lower in the two treated test areas as compared to their numbers in cotton fields in the two check test areas. The numbers in cotton in the treated areas were significantly lower in the last two weeks of July. Numbers of TPB found in cotton in the four test areas in 1998 were too low for statistical analyses.