The spatiotemporal patterns of evapotranspiration (ET) are the basic problem of the circulation of water resources. Several studies have reported the spatiotemporal patterns of ET. However, there is a lack of systematic analysis of seasonal ET patterns over large areas, and the factors affecting ET under variable environmental conditions are still uncertain. In this study, flux observation data was used to evaluate the performance of MOD16 (the MODIS global ET product) data under different conditions, and then the MOD16 data were used to analyze the interannual and seasonal variability of ET and its spatial distribution characteristics in China. Furthermore, path analysis was used to analyze the factors affecting the change of ET. The results can be summarized as follows: (1) MOD16 data are sufficiently reliable and meet the precision requirements for research; (2) the vegetation ET in China showed significant temporal and spatial differences at the interannual and seasonal timescales, respectively, with an average annual value of 359.61 mm between 2001 and 2018; (3) the ET in China showed an upward trend in 2001–2018, with an average change rate of 2.95 mm/year, and exhibited differences in various seasons and for different vegetation types; (4) surface vegetation are the main factors affecting ET in southeast of China, while in northwest, the ET variation is affected by multiple factors. Under the conditions of climate change and vegetation cover change, several strategies should be implemented to deal with the sustained increase of vegetation water consumption, for example, the adjustment of agricultural production structure and the optimization of planting structure. The ET change law obtained in this study can be used as a reference for water resource regulation and management.