<p>This study investigates the optimal design of truss structures at both macro- and nanoscale dimensions, focusing on the influence of size effects modeled through nonlocal elasticity theory. Recognizing the limitations of classical elasticity at small scales, particularly for nano-/microelectromechanical systems (NEMS/MEMS), this research employs the Artificial Bee Colony (ABC) algorithm to perform shape optimization of truss structures under frequency constraints. A MATLAB-based program was developed and validated using a 37-member macrotruss benchmark. The study then explored the impact of the atomic parameter (e₀a) on both macro- and nanoscale 37-member truss designs, as well as on 7- and 13-member nanoscale trusses from the literature. Findings indicate that while the ABC algorithm demonstrates robust performance in structural optimization, the significance of the atomic parameter varies with scale and design parameters. For macrostructures, increasing e₀a leads to a parabolic increase in optimal weight and shape changes, whereas the effect is less pronounced in nanostructures unless e₀a is scaled appropriately. Crucially, the optimization process consistently yielded lighter designs compared to unoptimized structures. The study concludes that considering size effects and employing metaheuristic optimization techniques like ABC are crucial for achieving efficient designs of truss structures, especially at the nanoscale, and highlights the need for further research into advanced material models to fully capture nanoscale behavior.</p>