Improvement of light penetration within tree canopies has been a constant objective of fruit tree architecture manipulation through the setting up of training systems. Recently, centrifugal training, i.e. the removal of fruiting shoots in the tree centre and on the underside of branches, has been proposed to improve fruit size and colour as well as return-bloom as compared to conventional solaxe-trained trees with equivalent crop loads. The present study was conducted to quantify the benefits of centrifugal training on light interception by the fruiting shoots via computer-assisted three-dimensional representations of foliage geometry. Data were collected on six 5-year-old apple trees cv.Galaxy, trained either with solaxe or centrifugal training systems, using an electromagnetic 3D digitiser. The 3D distribution of the foliage in the tree canopy was recreated by combining both the spatial locations of shoots (as measured from 3D digitising) and foliage reconstruction. Light interception efficiency properties of the trees were characterised by silhouette to total area ratio (STAR) values computed from images of the 3D mock-ups. Compared to the solaxe system, centrifugal training significantly improved the STAR of the whole tree by 20%. It also increased both leaf area and STAR of the fruiting shoots by approximately 15%, regardless of their position in the canopy. In this paper, we discuss the role of this enhanced light interception by the canopy in increasing the autonomy of the fruiting shoot, i.e. improved fruit size and colour, and return-bloom.