Human activities can heavily influence changes in net primary productivity (NPP), and for this reason, the quantitative calculation of NPP as influenced by humans (NPP_H) has a vital significance. In this paper, a model of human impact on NPP is proposed based on the coefficient of variation. The basic concept, theoretical foundation, and calculation are discussed, and the distribution and variation of NPP_H in the Shiyang River Basin are analyzed. The following conclusions were made. (1) The model is based on the indirect method, which avoids the complexities of human activity In addition, the NPP_H model has other advantages, including its scientificity in theory, fewer input parameters as it uses only the coefficient of variation, feasibility and easy obtainability technology, and a straightforward quantitative evaluation of human activity on NPP. (2) The effect of human activity on NPP was widespread and severe in the Shiyang River Basin during 2000-2010. The area with an annual average NPP_H value greater than 40 gC m^-2a^-1 was 96.21%, and the area with a degree of human impact greater than severe was 20%. There were two main features when the spatial distribution of the influence of human activities on NPP was considered. The first was the intense positive influence of human activity on three oases (Minqin, Wuwei, and Jinchang), where there is cultivated land, and the second is the strong negative influence on the junction of Sunan, Tianzhu, Gulang, and Yongchang, where the vegetation consists mainly of grassland and the predominant human-related activity is grazing. Thus, the human impact on NPP was both positive and negative, and significant for both. The average positive annual value as affected by humans was 1.63×10⁶ gC m^-2a^-1, and the negative value was 1.21×10⁶ gC m^-2a^-1, producing a net value of 4.20×10⁵ gC m^-2a^-1. The positive effect averaged 136.84 gC m^-2a^-1, and the corresponding negative effect was 100.32 gC m^-2a^-1; therefore, overall, the Shiyang River Basin was positively influenced by humans. The degree of human impact on different areas varied from slight to general, serious, and severe, with the percentages gradually decreasing from 45.96% to 27.11%, 18.58%, and 8.38%, respectively. The areas of positive and negative influence were roughly equal in degree, but the positive influence was slightly more than negative in the severe and general degree categories, and the negative influence was slightly more than the positive in the slight and serious degree areas. (3) Ganzhou county was the area most intensely and positively influenced by human activity. In contrast, Tianzhu county showed a strong negative influence. Minqin county showed a positive influence, which positive effect of the total in second place and effect of average in three place and negative effect all in second about total and average. The other areas influenced by human activity were Yongchang, Gulang, Sunan and Jinchang. (4) The NPP_H varied greatly in this period and showed as decreasing trend. In some regions, human influence has decreased (53.9%), while in other areas human activity has enhanced (46.1%). The positive effect was weakened by 8.12×10⁵ gC m^-2a^-1, the negative effect was weakened by 8.07×10⁵gC m^-2a^-1, and the positive effect was enhanced by 8.02×10⁵gC m^-2a^-1, and the negative effect was enhanced by 3.94×10⁵gC m^-2a^-1, giving a net reduction of 4.25×10⁵gC m^-2a^-1, indicating that overall, human activity showed a decreasing trend.