A simplified method for building three-dimensional (3D) mock-ups of peach trees is presented. The method combines partial digitizing of tree structure with reconstruction rules for non-digitized organs. Reconstruction was applied at two scales: leaves on current-year shoots (CYS) and shoots on 1-year-old shoots (OYOS). Reconstruction rules make use of allometric relationships, random sampling of shoot attribute distribution and additional hypotheses (e.g., constant internode length). The method was quantitatively assessed for two training systems (tight goblet and wide-double-Y), at a range of spatial scales. For this purpose, light interception properties of reference and reconstructed mock-ups were compared. Mock-up quality depended on scale. Foliage reconstruction on CYS was unsuitable for generating a given CYS. Similarly, CYS reconstruction on OYOS was unsuitable for generating a given OYOS. This is because generic rules derived at the population scale do not consider specific foliage or shoot attributes of a given CYS or OYOS. In contrast, foliage reconstruction on CYS was able to generate OYOS mock-ups having light properties similar to the reference mock-ups. The same held for CYS reconstruction on OYOS for light capture properties at the tree scale. The CYS reconstruction on OYOS was also suitable for deriving OYOS distribution as a function of light interception ability. Reconstruction rules were successfully used to build the vegetation neighborhood of a reference shoot. The proposed method could therefore be used to make 3D tree mock-ups usable for a range of some, but not all, light computations. Because the simplified method allows large time savings, it could be used in virtual experiments requiring large numbers of replicates, such as comparative studies of tree genotypes or training systems.