Abstract: To explore spatial structure of secondary forest communities at different succession series in the hilly area of northern subtropical China, we set up 36 sampling plots in the Lishui Foresty Farm, Jiangsu Province, and conducted detailed field investigations on the woody plants in these plots. We distinguished four typical community types using canonical correlation analysis(CCA)across our study plots, and then used point pattern analysis to characterize the spatial distribution patterns of the main tree species for these community types. Our results suggested that in the study area typical forest communities were represented as Pinus massoniana-Quercus acutissima community, Ilex chinensis-P. massoniana community, Q. acutissima-Q. variabilis community and Cunninghamia lanceolata-Vernicia fordii community. Soil conditions appeared to be the primary factor resulting in the difference of community composition. Most dominant tree species in the study communities showed uniform patterns on the scale of >3 m, suggesting that intraspecific competition underlies such pattern formation. Interspecific spatial association often manifested neutral patterns, except for those species such as Q. acutissima and Q. variabilis, which presented significant negative correlations. Interspecific competition due to similar functional traits of these tree species was a likely driving mechanism of such association. Our study could provide a useful reference to effective management of secondary forests in northern subtropical China.