An action threshold (AT) based on the specific pest–crop relationship is an important element of integrated pest management (IPM) programs. We evaluated planting time-based ATs to manage a sucking pest, cotton leafhopper (Amrasca devastans; Hemiptera: Cicadellidae) in conventional (CIM-554) and transgenic (Bt.CIM-599) cotton cultivars under field conditions. Cultivars were planted on 15 March, April, and May during 2011 and 2012. Only two ATs (0.1 and 1.0 A. devastans per leaf) and an untreated control could be tested on 15 March planted cotton, as the populations of A. devastans never reached a density of two per leaf. Cotton planted on 15 March had the highest seed cotton yield, highest net and marginal rate of returns (MRRs) than other planting times. At this planting time, only one insecticidal application was needed at an AT of 1.0 A. devastans per leaf, with no significant yield loss. Four ATs (0.1, 1.0, 2.0, and 3.0 A. devastans per leaf) and an untreated control were tested for cotton planted on 15 April and 15 May. For the 15 April planting, 2.0 AT led to 3 applications as compared with 10 applications needed with 0.1 AT; without any significant yield loss, generating the highest net return and MRR for this planting date. The 15 May planting was more vulnerable to A. devastans damage and 1.0 AT generated the highest net return and MRR with four applications as compared with all other ATs.These results suggest that the use of planting time-based ATs can improve cotton yield through cost-effective applications of insecticides.