The aim of the project is to control the failure of the structural system of an electric transmission tower system during strong earthquakes. Optimization process is being applied by the support of both surrogate modeling and Hessian Matrix method using MATLAB codes. Latin Hypercube sampling method is dedicated to arrange twenty-five numerical models of the system in ABAQUS program. Five parameters are used in the optimization process which are the thickness of steel member of the transmission tower, the Modulus of Elasticity of the steel member, radius of the conductor steel wire, Modulus of Elasticity of the conductor steel wire, and the Peak ground acceleration of the earthquake. The first four parameters are controllable in the design but the last parameter is can't be controlled because it is uncertain and it is related to the earthquake strength. The surrogate models for the maximum principal stress and maximum principal strain of the structural system are constructed. The coefficients of determination of both surrogate models were 97.02% and 92.06% respectively, which are considered excellent and reliable for representing and predicting the responses of the structural system. The results of the optimization all involved five parameters from the Hessian Matrix method for both surrogate models were very accurate and are optimum values through comparing the responses of them in the surrogate models with the numerical simulation results. The surrogate modeling and the Hessian Matrix method are great tools which can be utilized in the optimization of the design of the electric transmission tower system under strong earthquakes. Consequently, the failure of the structural system then can be easily controlled to a great extent in the design stage and after construction.