This study took the 4 hm² sample plot as the research object in Guandi Mountains, Shanxi Province, and trees were continuously positioned and monitored according to Center for Tropical Forest Science (CTFS) technical specifications. Based on the observation data of tree radial growth in the sample plot from 2010 to 2015, combined with the survey and sampling measurement data of environmental factors such as terrain variables and soil variables, we analyzed the spatial autocorrelation patterns of radial growth of tree populations and their variation with different habitat types, and explored the influencing factors of the radial growth of the tree populations. The results showed that Picea wilsonii, Larix principis-rupprechtii, Betula albosinensis, Betula platyphylla, and Quercus wutaishanica were main tree species in the secondary Picea forest, which were distributed in four habitat types (ridge habitat, low-altitude gentle slope habitat, high-altitude gentle slope habitat, and low-lying steep slope habitat) with different diameter structures. The mark correlation function (MCF) analysis showed that the spatial autocorrelation characteristics of radial growth of the five tree species were different in the same habitat type. For the same tree species, the spatial autocorrelation characteristics of radial growth was not only scale-dependent, but also changed with different habitat types. The linear mixed-effects model (LMM) analysis showed that the significantly positive effect of initial size, i.e diameter at breast height (DBH), on tree radial growth was prevalent in all populations of all habitat types in the plot. The significant effects of biotic factors on radial growth of trees were detected only in the population of Picea wilsonii in specific habitat types, indicating that the radial growth of trees was affected by conspecific neighbors, but the significance of its influence varied with tree species. The environmental factors had significant effects on radial growth of specific tree species in specific habitat types. Among the environmental factors, altitude and convexity had significantly negative effects on radial growth, and slope, soil principal coordinate analysis (PCA) axis 1 (acidic pH and higher content available manganese, available nickel, available iron, available copper), and soil PCA axis 2 (higher content of total salt, available nitrogen, available potassium, organic matter) had significantly positive effects on radial growth. The radial growth of trees and the formation of spatial autocorrelation patterns of trees in secondary Picea forest were jointly driven by the initial DBH, neighbors, and habitat factors.