The fraction of absorbed photosynthetically active radiation (FPAR) is a key biophysical variable for vegetation productivity estimation, vegetation growth condition monitoring, and climate change analysis. The object of this study is to conduct a comprehensive sensitivity analysis on the contributing factors of FPAR variation, in order to reveal the impacts of the interference factors on FPAR estimation using remote sensing and to improve the accuracy of the model. The analysis was conducted based on a coupled leaf-canopy–atmosphere radiative transfer model and a global sensitivity analysis algorithm. Due to the impact of diffuse FPAR on the variation of FPAR, we also analyzed the determinant factors of diffuse FPAR. The results showed that leaf area index (LAI) and the average leaf inclination angle (ALA) are two important canopy structural parameters determining FPAR variability, especially before canopy closure. FPAR was found to be affected differently by these two variables with canopy development. Canopy background reflectance has a great impact on FPAR estimation at low LAI values (LAI<1.0). For the effects of leaf biochemical variables, the contribution of leaf chlorophyll concentration (LCC) to FPAR variability is negligible at the early vegetative growth stage, but becomes dominant starting from the exuberant growth periods, with the contribution increasing consistently with LAI. For the external factors, solar zenith angle (SZA) and aerosol optical depth (AOD) are two important driving factors significantly influencing FPAR by changing the optical path length and the proportion of the diffuse photosynthetically active radiation (PAR). SZA influences both FPAR and diffuse FPAR throughout the growing period, while, AOD is the most important determinant factor for diffuse FPAR variation. When LAI is low, the impact of canopy background reflectance on the diffuse FPAR cannot be ignored. The analysis also revealed that the diffuse FPAR variation significantly influences on the vegetation productivity estimation. In summary, this study shows that the variation of FPAR is affected by different factors in different ways. It provides a comprehensive understanding of the contributing factors to FPAR variability including the diffuse FPAR, which could be potentially useful in developing FPAR retrieval algorithm from remote sensing data.