In 2006, two replicated “regressive spray” test were conducted in NC.  In 2007, five replicated “regressive spray” test were conducted with two in NC, two in GA, and one in SC.  Previous data suggests that protecting cotton during the first three weeks of bloom from stink bugs has little impact on yields.  However, protecting cotton four and five weeks post bloom has a significant impact on yield.  The purpose of this study is to help determine the impact of stink bug feeding on yield later in the cotton growing season.  In order to minimize impact of caterpillar damage, tests were conducted in Bollgard II varieties.  Each test consisted of six to eight rows by 50 to 100 feet with four replications.  Spray treatments consisting of a high rate of dicrotophos (Bidrin 8E @ 0.5 lb. ai/acre) plus the highest rate of a pyrethroid.  Treatment sprays began at anthesis and continued throughout the growing season for an average of seven spays in the most protected plots.  Each week, one treatment was removed from the spay schedule until only the most protected plot remained.  Weekly data collections for most test sites included square retention, dirty blooms, beet sheet samples, internal boll damage to quarter-sized bolls, damage to bolls just prior to harvest, boll diameters, yield and quality.  This data could have implications upon developing and implementing a dynamic threshold in areas where stink bugs are the predominate pest.  A higher level of protection against stink bugs could occur during a two week period four and five weeks post anthesis, while allowing for lower protection early and later in the growing season.