The Enshi area is the largest tobacco growing base in Hubei Province, China, and the peach aphid, Myzus percicae (Sulzer), is one of the most important pests of tobacco leaves in this region. The prevalence of peach aphid populations is increasing rapidly, as are incidents of viral disease spread by these insect pests, both of which result in substantial losses to tobacco production. More research on the spatial distribution and development of peach aphids will enhance the ability to predict its occurrence and comprehensively manage its populations. The alate form of the aphid migrates between tobacco fields and areas, whereas the population dynamics and occurrence cycles depend more on the insect's apterous type. Because the spatial structure of peach aphid populations in tobacco fields depends on the host's physiological and biochemical characteristics at different growth stages, more research is required to fully understand it. Classical statistical methods assume a random distribution, but insect populations show strong spatial correlation in the field. Newly-developed geological statistical methods can accommodate spatial correlation, allowing analyses of the spatial distribution of regionalized variables. We investigated the density of apterous peach aphids in tobacco fields at the seedling, rosette, rapid-growth, and maturity stages of growth and used geostatistical methods to analyze the spatial features and development of the insect, to simulate its distribution at different stages of tobacco, and to analyze the correlation between its spatial distribution during different stages of tobacco. The results showed that the aphid density was lowest at the tobacco seedling stage ((5.59 ± 4.07) aphids per plant) and peaked at the rapid-growth stage ((14.5 ± 9.6) aphids per plant). The variable coefficient of the density was relatively large (0.6147-0.7281), which indicated the inhomogeneity of the spatial distribution. The coefficient decreased as the population density increased, which showed that the population structure stabilized at higher densities. The largest kurtosis in the spatial pattern occurred at the seedling stage of tobacco, indicating more spatial clustering. We divided the plane into four directions (0°, 45°, 90°, and 135°) and found that the spatial distribution of aphids in the 135° direction during the seedling stage and the 45° direction during the rosette stage were randomly distributed. In the 0° and 45° directions during the period, the distributions matched linear models with a base station, and all other distributions fitted nested models with spherical + exponential factors. Therefore, we deduced that the primary model for apterous M. persicae was the latter, indicating an aggregated distribution. All nuggets, stills, and ranges increased with density. The random degree of the model in the seedling stage (0.1905-0.7186) was significantly greater than in other stages (0.0116-0.1620). Kriging interpolation clearly described the migration in the seedling stage and showed that the spatial distribution was relatively stable in the other three stages. The spatial pattern of apterous peach aphid in the seedling stage of tobacco had no correlation with that in the rosette stage, but there was an obvious correlation between the rosette and rapid-growth stages and also between the rapid-growth and maturity stages. This indicated that migration should occur during the seedling stage and that the population distribution should remain relatively stable afterwards. We demonstrated that dividing the plane coordinate into four directions made the description of the spatial distribution more accurate and comprehensive.