The study of vegetation cover change is an important tool to evaluate the change of terrestrial ecosystem, but the previous studies mostly used linear trend to express the change of vegetation cover, ignoring the change of trend over time. Using GLASS FVC data, BFAST method and landscape pattern analysis, the nonlinear trend and distribution pattern of vegetation cover change in China were explored from 1982 to 2018. The results showed that:(1) compared with the results of linear trend method, although the vegetation cover in Sichuan Basin and the Loess Plateau presented a significantly linear increase trend, the detection results of BFAST showed that the increase trend was interrupted. Meanwhile, in Qinghai and Northeast China, the vegetation cover experienced a shift from degradation to improvement rather than simple linear increase, while it changed from the first improvement trend to later degradation in the middle east of Qinghai-Tibet Plateau. (2) Among the nonlinear trend types, the monotonic increase trend accounted for 33.58%, mainly distributed in the Inner Mongolia, Shaanxi, Henan and other places; the monotonic decrease trend accounted for 1.82%, mainly distributed in the southeast coastal area; the interrupted increase trend accounted for 22.91%, mainly distributed in the east of Sichuan Basin and the North China; the interrupted decrease trend accounted for 2.68%, mainly distributed in the southeast of Qinghai Tibet Plateau; the trend shifted from increase to decrease type accounted for 4.20%, mainly distributed in Qinghai; the trend shifted from the decrease to increase type accounted for 14.62%, mainly distributed in Jilin. The large-scale increasing trend of vegetation cover fully reflected the improvement of vegetation in China during the past few decades, but the monotonic decrease and the increase to decrease trends showed that the potential risk of vegetation degradation could not be ignored. (3) The time of change of different trend types were different, less occurred from 1988 to 1999, but mainly occurred from 2000 to 2011, which has been probably affected by the large-scale ecological protection and restoration projects since the 21st century. (4) The patches of the vegetation improvement trend type (monotonic increase, interrupted increase, decrease to increase) were characterized by large aggregation, small dispersion, and complex shape driven by strong interference, indicating that the process of vegetation cover change was actually very complex with high diversity on the regional scale; the patches of degradation trend type were small and dispersedly distributed. This study shows that the nonlinear trend and pattern analysis can more accurately evaluate vegetation cover change, so as to provide a scientific reference for the protection and recovery of ecological environment.