The intrinsic defects in lithium niobate (LiNbO3) bulk crystals have been investigated with experimental and theoretical methods for many years. According to the most widely accepted model for the intrinsic defects in LiNbO3 bulk crystals, antisite niobium (NbLi) and lithium vacancy (VLi) defects are present in the lattice in concentrations depending on the composition, i.e., the Li/Nb ratio of the crystal. But nowadays, scientific attention has turned to nanocrystals. One of the methods to produce nanosized LiNbO3 crystals is the frequently used ball-milling procedure. In the present work, sintered lithium niobate with congruent composition was ground under wet conditions in a high-energy ball mill for different grinding times. After that, thermal treatments were performed on the powder samples at temperatures of 300 and 500 °C. The structural changes were monitored by Raman spectroscopic measurements. Atomistic computer simulations based on classical potential models have long been used for understanding intrinsic defect formation in various crystal lattices. In this work, a Li245Nb245O735 unit, constructed from a 5 × 5 × 5 supercell with sharp corners cut away to achieve a size of about 24 × 24 × 52 Å3, was chosen for GULP force–field calculations as the starting configuration. The relaxation of the unit was achieved by first using the conjugate gradient and then the Newton–Raphson optimization methods. This process was repeated for the relaxed nano-LiNbO3 structure by placing one or two NbLi + 4 VLi, or one LiNb3O8-type neutral intrinsic defect into the supercell. The total energies were compared in terms of both the defect positions and their relative locations.