Truss structures are common in the building industry. One way to contain construction costs is to implement structural optimization. Optimization has to consider cross-sectional size, area, topology, and node coordinates as design variables. However, each truss structure has numerous complex constraints and variables that make optimizing this structure complex and difficult. The metaheuristic method is efficient and effective in solving large and complex problems. This paper tested three metaheuristic algorithms: particle swarm optimization (PSO), differential evolution (DE), and symbiotic organisms search (SOS). Each algorithm was used to optimize a 10-bar planar truss structure and a 15-bar planar truss structure. SOS was found to have the best optimization results, convergence behavior, and consistency.