Full-season interference and critical duration of early-season ivyleaf morningglory interference with cotton were measured in experiments at two Oklahoma locations. Densities ranging from 0 to 64 weeds/10 m of row were evaluated for their effects on cotton lint yield, fiber quality, and harvest efficiency. Analyses using linear-linear regression proved superior to linear, curvilinear, and linear-plateau models. Lint yield reductions of 36.9 kg/ha (or 5.9%) were recorded at Perkins for each weed/10 m up to 8.7. At densities greater than 8.7/10 m, yield was reduced an additional 3.0 kg/ha (0.5%) for each additional weed. At Chickasha, lint yield was reduced 29.7 kg/ha (or 3.9%) for each weed/10 m up to 9.0 with an additional lint loss of 3.6 kg/ha (0.7%) for each weed above the density of 9.0/10 m. Yield losses were over 80% before fiber length was also reduced. Mechanical harvest was not possible at densities above 8 weeds/10 m at Chickasha and above 16 weeds/10 m at Perkins. In critical duration experiments conducted at Perkins, lint yield was reduced 52.9 kg/ha (or 11.2%) for each week weed removal was delayed up to 9.5 wk. An additional 1.0 kg/ha (0.2%) was lost for each week of interference after 9.5 wk. At Chickasha, lint yield was reduced 49.0 kg/ha (or 7.8%) for each week of interference up to 11.0 wk with an additional loss of 1.2 kg/ha (0.2%) for each week thereafter.