Crop managers need to determine the most profitable time to defoliate cotton (Gossypium hirsutum L.) in a high rainfall environment such as the coastal region of Texas. In cotton production, delaying defoliation exposes open bolls to a higher probability of rainfall, and thus, reduces lint yield and fiber quality. Premature defoliation, however, has detrimental affects on lint yield and fiber quality. A more recent method to determine defoliation is based on heat-unit (HU or DD15) accumulation after physiological cutout or five nodes above white flower (NAWF=5). Results have been inconsistent across a wide range of field environments when utilizing HU accumulation past cutout; therefore, adoption of this method has been limited. Many regions of the Cotton Belt have maximum day time temperatures during the growing season that are above optimum for maximum growth. Field studies were conducted for three consecutive growing seasons in the Brazos River Valley and Upper Gulf Coast regions of Texas. The purpose of this research was to identify an upper temperature threshold (UTT) for calculating degree days for defoliation timing. The experimental design consisted of a split-plot design with four replications. The main plots consisted of three upper temperature thresholds (32°C, 35°C, and no upper limit) and the subplots were five HU timings (361, 417, 472, 528, and 583) accumulated from date of cutout. Utilizing an UTT to calculate daily HU failed to explain differences in the optimum time to defoliate based on accumulated HU from cutout for the upper thresholds investigated. Accumulated HU had a significant impact, however, on defoliation timing. Comparison of the two locations showed that maximum lint yield was obtained at 472 HU and 52% open boll at Wharton County versus a maximum of 528 HU and 62% open boll for the Burleson County location. Employing the NACB=4 method to time defoliation at both locations would have resulted in premature application of harvest aids and reduced lint yields. No differences were observed in adjusted gross income values at Wharton County among the 417, 472, 528, and 583 HU treatments. For Burleson County, adjusted gross income peaked in value at 528 HU.