To complement the current buffer determination studies that do not distinguish between differences in employees' capabilities, and to adequately consider the influence of human resources on buffer sizing, a buffer sizing method that considers employees' capabilities attributes is proposed. Firstly, the formulae of four employee subjective and objective capabilities factors, namely, procrastination degree, work intensity, multitasking capability and skill level, are proposed; secondly, the entropy-weight-TOPSIS method is used to determine the weights of the four indicators and calculate the capabilities attribute coefficients of employees on each activity; then, the activity safety time is extracted based on the capabilities attribute coefficients and the buffer size is determined based on the central limit theorem; finally, Monte Carlo simulation is conducted. Comparing the proposed method with the cut and paste method and the root square error method, the simulation results show that the proposed method can improve the project completion rate while reducing the buffer size and decreasing the project duration.