Optimization of truss structures under technical constraints has always been one of the fundamental challenges in structural engineering due to the vastness of the search space and the complexity of relationships among design variables, requiring high precision in navigating computational paths. This paper introduces the application of a novel metaheuristic algorithm called Catch Fish Optimization (CFOA), inspired by the collective intelligence and strategic behavior of fishermen, offering an efficient solution to overcome the limitations of classical methods and achieve accurate answers. The core innovation of this algorithm lies in establishing a dynamic balance between exploring new regions and exploiting favorable points, which not only ensures stability but also significantly increases the convergence speed. For the first time in solving optimization problems of truss structures the performance evaluation of CFOA has been conducted on reference models of 10-member, 18-member 2D, and 25-member spatial trusses. Specifically, this algorithm has succeeded in improving the structural weight in the aforementioned models by 0.0004%, 1.35%, and 0.00046%, respectively, compared to the best previous results. Although these improvements appear small in some cases, they are of great importance in structural optimization problems with very tight tolerances. Ultimately, this algorithm has proven its high capability in finding the most economical and stable engineering solutions in complex problems as a powerful and reliable tool.