Abstract: In order to better understand the occurrence regularity of Myzus persicae (Sulzer) (Hemiptera:aphidiae) and to provide theory evidence to the method that the natural enemies are used to control Myzus persicae in Liangshan tobacco regions of Sichuan Province, two regression modelsIwao's regression and Taylor's power principle and eight indices of aggregationK, C, C_A, I, M^*, m^*/m, L^*, L^*/(1+m) were used to test the spatial distribution patterns of Myzus persicae population. The results indicated that the distribution patterns of Myzus persicae population, which accords with minus binomial distribution belong to an aggregated distribution. The aggregated mean indice(λ) suggested that aggregation was caused by biological characteristics of Myzus persicae and environmental factors. In addition, the aggregation indexsK, C_A, m^*/m, L^*/(1+m) can explain the spatial distribution patterns of Myzus persicae population better. The equation of theoretical sampling number and sequential sampling model were inferred from the sptial pattern parameters. In conclusion, the distribution patterns of Myzus persicae population was an aggregated distribution and their aggregation seemed to be due to both the biological characters of population and environmental factors.