Due to the complexity and uncertainty of affecting factors on the distribution characteristics of wind loads on the surface of structures, it is difficult to fully obtain and describe their wind load distribution characteristics solely through wind tunnel tests. Therefore, an inquiry has thus been made into the prediction of structural surface wind pressure distribution based on network function interpolation theory. Firstly, with the introduction of a network function interpolation method with multi-directional fitting and one-time error adjustment, based on the network function interpolation technology fused with data obtained from multiple sensors, a method has been proposed for the prediction of the wind pressure distribution on the surface of a structure, which is validated by using wind tunnel test data. The example verification results show that the network function interpolation method can quickly calculate the surface wind pressure values of adjacent positions based on the measurement point network arranged on the surface of the structure, making it an effective way for the prediction of the surface wind pressure distribution characteristics of the structure as a data fusion method. Moreover, this proposed method is characterized with such advantages as fine stability, simple practicality, high accuracy, applicable to reducing the number of measurement points and expanding the wind pressure distribution data of complex building skins in wind tunnel tests.