Abstract: The low-latitude plateau region in China features a rich variety of surface types and is an important carbon sink area sensitive to climate change. However, owing to the scarcity of direct terrestrial net ecosystem exchange (NEE) observations in this area, an in-depth understanding of the carbon sink status is lacking. This study utilized four sets of NEE data from different sources to comparatively analyze the spatial and temporal distribution differences of NEE in the low-latitude plateau region of China. The relative uncertainty between different data sets was assessed using the generalized triangular hat method, with the aim of providing consistent data to guide studies of terrestrial ecosystem carbon sinks in this region. The results of the study indicated the following: in terms of the annual average values of the four sets of NEE data, GCDF (Global Carbon Fluxes Dataset), FLUXCOM, and GCAS (Global Carbon Assimilation System) all exhibited carbon sinks with values of −423.23, −321.25, and −63.57 g(C) m⁻² a⁻¹, respectively, while Ms TMIP (Multi-scale Synthesis and Terrestrial Model Intercomparison Project) data exhibited a carbon source with a value of 2.63 g(C) m⁻² a⁻¹. In addition, all four data sets showed negative values during summer and autumn, indicating a clear carbon sink; FLUXCOM and GCDF data showed carbon sinks during spring and winter, while GCAS and Ms TMIP data showed carbon sources. In terms of spatial distribution, all four sets of data exhibited carbon sources in winter and carbon sinks in summer across the entire low-latitude plateau region. Furthermore, Ms TMIP data showed a strong carbon source in the Yunnan region in spring, while GCFD data showed a strong carbon sink in summer. In terms of different vegetation types, FLUXCOM and GCDF data showed carbon sinks in all seasons except for grasslands, and the annual average values of carbon sinks were most significant under deciduous broadleaf forests, except for Ms TMIP data. Finally, in terms of relative uncertainty, the relative uncertainty of the FLUXCOM data was the smallest of the four data sets in all seasons and throughout the year.