Abstract: Cyrtorrhinus livdipennis (Reute) is one of natural enemies of Nilaparvata lugens (Stål), which more importantly has a clear accompanying migration pattern. The study of the effects of irrigation method on the in-and-out migration of N. lugens and C. livdipennis of high-quality late rice field could provide theoretical and technical support for an integrated control of N. lugens and reduce the use of chemical pesticides in cultivation. This experiment designed two isolation (semi-isolation and fully-isolation) modes to study the effects of deficit irrigation, wet irrigation, intermittent irrigation and long-term irrigation on immigration of N. lugens and C. livdipennis in high-quality late rice fields in 2014-2015. The results showed a significant difference on immigration and emigration of N. lugens and C. livdipennis in high-quality late rice fields between two years. There was no N. lugens and C. livdipennis immigration under all irrigation treatments in 2015 due to low temperatures and frequent rainfalls during mid to late growth stages of late rice. However, with higher temperatures in 2014, there was a significant migration. Based on the state of immigration, N. lugens immigration was lowest under deficit irrigation, which occurred early with a large base number, and was only 0.6% of that of intermittent irrigation. The intermittent irrigation had the largest number of immigration. Immigration rate (immigration number divided by increase in immigration number) of N. lugens under intermittent irrigation was 19.4, 1.4 and 3.4 times that of deficit, wet and long-term irrigation, respectively. The time of C. livdipennis immigration under deficit irrigation and wet irrigation was 8 days earlier than that under the other methods of irrigation. Based on the state of emigration, the time of N. lugens emigration under long-term irrigation was 11 days earlier than the other modes of irrigation. The largest number of N. lugens emigration under intermittent irrigation was 1.4, 1.1 and 5.8 times that of deficit, wet and long-term irrigation, respectively. The emigration rate (emigration number divided by decrease in emigration number) under wet irrigation was 1.3, 1.1 and 1.8 times that of deficit, intermittent and long-term irrigation, respectively. The number of C. livdipennis emigration in deficit irrigation increased by 102.9%, 127.2% and 1709.1% compared with wet, intermittent and long-term irrigation whereas emigration rate increased by 7.0%, 83.4% and 49.7%, respectively. Therefore, deficit irrigation reduced immigration of N. lugens, but increased the population of N. lugens before immigration compared with other irrigation methods. It simultaneously promoted emigration of enemy C. livdipennis, which increased the risk of N. lugens outbreak and control difficulty of N. lugens. Long-term irrigation effectively prevented N. lugens from immigration and enemy C. livdipennis from emigration. Wet and intermittent irrigation simultaneously promoted N. lugens emigration, while intermittent irrigation significantly reduced emigration of enemy C. livdipennis (P < 0.05).