A new method for optimizing the design of nonlinear sinusoidal-web steel beams is introduced in this paper, focusing on reliability-based design principles. Utilizing a custom-written code with a reliability index as a key control factor, the study incorporates stochastic variables such as flange thickness, sinus wave width, sinusoidal-web plate thickness, and applied load magnitudes. Employing Finite Element Analysis (FEA) through ABAQUS software conducting experimental testing on sinusoidal HEA beams, this research showcases the successful application of reliability-based design in modifying beam design to meet the reliability requirements. The adoption of the suggested approach, which incorporates Monte Carlo simulation, is crucial in guiding the development of structural configurations that are robust to random nature in both load and manufacturing conditions. Results showcase improved structural integrity, and the successful convergence of optimized values, emphasizing the potential for enhanced steel structure design.