Based on moderate resolution imaging spectroradiometer (MODIS) normalized difference vegetation index (NDVI) remote sensing data, the fractional vegetation cover in the upper reaches of Minjiang River was estimated using a pixel dichotomy model. Furthermore, the temporal and spatial variation pattern and stability of the fractional vegetation cover in the upper reaches of Minjiang River from 2006 to 2016 were studied using univariate linear regression and stability analyses. The damage degree of the vegetation in the upper reaches of Minjiang River and the vegetation restoration after the earthquake are discussed in sections. The influencing factors of fractional vegetation cover in the upper reaches of Minjiang River were detected using the Geographic detector model. The driving force of vegetation cover change in the upper reaches of Minjiang River was analyzed. The results showed the following. (1) The fractional vegetation cover in the upper reaches of Minjiang River from 2006 to 2016 was good, and the overall fractional vegetation cover was stable. The annual average fractional vegetation cover was 0.79, and the area with fractional vegetation cover > 0.8 accounted for 69% of the total area of the upper reaches of the Minjiang River. (2) The "5.12 Wenchuan earthquake" in 2008 seriously damaged the vegetation in the upper reaches of Minjiang River. The degraded area of fractional vegetation cover was 14013.41 km², which accounted for 57% of the total area of the upper reaches of the Minjiang River.The vegetation restoration condition in the upper reaches of Minjiang River from 2008 to 2016 was good.The area of fractional vegetation cover improvement was 17390.69 km², which accounted for 71% of the total area of the upper reaches of the Minjiang River. The fractional vegetation cover in the upper reaches of Minjiang River exceeded the pre-earthquake level. (3) The fractional vegetation cover in the upper reaches of Minjiang River was mainly affected by four factors:altitude, temperature, agrotype, and precipitation. Its explanatory power was > 40%, the interpretation of the Geomorphic type and vegetation types was between 20% and 40%, and the interpretation of slope and aspect was < 1%.