Net primary productivity (NPP) and net ecosystem productivity (NEP) are important indicators for estimating the carbon source/sink of terrestrial ecosystems. NEP is the net carbon absorption of vegetation ecosystems, which can quantitatively describe the carbon source/sink capacity of vegetation ecosystems. It is great significance to analyze the spatiotemporal variation characteristics of NPP and NEP for scientific assessment of carbon source/sink function of terrestrial ecosystem and carbon emission trading, because Yunnan is one of the main regions of carbon sink in China. Based on the boreal ecosystem productivity simulator (BEPS) of gross primary productivity(GPP), NPP and NEP products from 1981 to 2019, the spatiotemporal variation characteristics of NPP and NEP and their applicability in Yunnan were studied by using linear trend analysis, literature comparison, and other methods. The results show that: (1) the NPP and NEP in Yunnan fluctuated horizontally from 1981 to 1999, and increased significantly after 2000. The average values of NPP and NEP per unit area from 2000 to 2019 were 739.0 gC/m² and 70.1 gC/m², respectively, with a total amount of 0.291 Pg C/a and 0.028 Pg C/a. Compared to 1981 to 1999, it increased by 12.4% and 150.4% respectively. The high value areas of NPP were mainly distributed in the west and south, most regions had an annual average NPP of over 700 gC/m², while the NEP were mainly distributed in the eastern and partial western from 2000 to 2019. (2) During the 2000 to 2019, most areas of Yunnan NPP and NEP showed an upward trend, except for the northwest, where NPP showed a downward trend. (3) The peak value of NPP occurred in July or August, and the valley value occurred in February. The month of NEP peak occurrence was basically the same as NPP, but the valley month was delayed 1-3 months. June to October were the main months of carbon sink. (4) The NPP simulated by the BEPS model was consistent with the widely used Carnegie-Ames-Stanford Approach (CASA) and remote sensing models, and the spatiotemporal change characteristics were conformed with the ecological restoration measures and climate characteristics of Yunnan. The simulated NEP was same as the integrated biosphere simulator (IBIS) and the carbon and water flux observation data of ChinaFLUX. Most of the spatiotemporal changes ware also basically consistent with the ecological restoration measures and climate characteristics of Yunnan, indicating that the BEPS model had good applicability in Yunnan.