[Objectives]  To identify scientific, accurate, and effective, sampling methods for estimating the population density of Lygus pratensis in cotton fields, and provide a theoretical basis and guidance for the prevention and control of L. pratensis in Xinjiang. [Methods]  Three sampling methods were used to estimate the population density of L. pratensis in cotton fields during 2015-2016 and the results compared. [Results]  (1) The sweep method detected peak periods of adult abundance one week earlier than the overall visual and partial visual methods. However, the overall visual method detected the occurrence and peak abundance of nymphs one week earlier than the sweep method. The trend in nymph abundance was consistent between the overall visual and partial visual methods. (2) There were significant differences between the different methods used to assess population density. The population density of nymphs and adults estimated by the overall visual method was significantly higher than that obtained by the sweep and partial visual methods. The population density of adults estimated by the sweep method was higher than that obtained using the overall visual method. [Conclusion]  The sweep method can accurately reflect the occurrence of adults but the overall visual method is more suitable for nymphs. Practical applications using a combination of the overall visual and sweep methods can aid the accurate forecasting and control of L. pratensis.