Estimating forest variables, such as photosynthetic light use efficiency, from satellite reflectance data requires understanding the contribution of photosynthetic vegetation (PV) and nonphotosynthetic vegetation (NPV). The fractions of PV and NPV present in vegetation reflectance data are typically controlled by the canopy structure and the respective viewing angle. The persistent but highly varying anisotropic behaviour of the forest canopy implies that there is canopy structural information to be exploited from multi-view angles measurements. In this work, a combination of radiative transfer modelling (FLIGHT) and linear unmixing techniques were used to isolate angular PV and NPV fractions from multi-angular CHRIS-PROBA (Compact High Resolution Imaging Spectrometer-Project for On-board Autonomy) data in order to assess their effects on a suite of vegetation indices. Angular variability in the NIR wavelengths contributed most to the angular change in PV and NPV fractions. In turn, for those pixels where the NPV fractions from near-nadir to backscatter were increasing, moderate correlations were found with the angular variability of the calculated vegetation indices. From these fractions, a Normalized Difference NPV Index (NDNPVI) was developed as a proxy for volumetric canopy composition.