Abstract: The maximization of stand spatial structure optimization of public welfare plantations achieved with minimal human intervention contributes to the enhancement of forest ecosystem service functions. Under the predefined threshold (where the harvest volume does not exceed 15%), seven spatial structure parameters, i.e. uniform angle index (W), species mingling (M), crowding (C), diameter dominance (U), opening degree (K), competition index (CI) and storey index (S), were selected on the basis of field data from three plots (30 m × 30 m) of Pinus massoniana. Based on multiplication and division method, spatial structure unit evaluation indices were constructed to evaluate the growth potential of forest so as to determine the trees to be harvested. A comprehensive evaluation indices for the stand spatial structure was established to evaluate the state of the structure, with the aim of determining the optimal number of trees for harvesting. The results showed that plots 1, 2 and 3 contained 112, 95 and 133 trees, respectively, while the optimal number of trees to be harvested were 21, 16 and 23, respectively. Notably, the volumes corresponding with the optimal number in plots 2 and 3 were below the harvesting threshold. The comprehensive stand spatial structure index (TF) increased by 15.4%, 14.3%, and 16.7%, respectively, indicating an obvious effect of improvement. This study proposes an optimization method for the spatial structure of P. massoniana public welfare plantations stands. Though computer-simulated logging, the maximum spatial structure optimization with minimal human intervention is achieved under the condition of the harvesting threshold, eventually offering practical value for forest management.