Protection from windblown sand is one of the key engineering issues for construction and maintenance of human infrastructures in arid environments. In the last decades, a number of barrier-type Sand Mitigation Measures with different shapes have been proposed in order to overcome this problem. Sand barriers are often deployed alongside long line-like infrastructures crossing vast desert regions. It follows that highly optimized preliminary design of the barrier cross section is of paramount importance in the perspective of a large-scale production, in order to minimize the construction costs per unit length, and maximize the aerodynamic performances. The present computational study aims to adapt and apply aerodynamic optimization to a windblown sand barrier. The search for the optimum is carried out on Computational Fluid Dynamics simulations, without recourse to surrogate models. Both gradient-based method and genetic algorithm are used, in the light of the features of the goal function previously sampled by extensive sensitivity studies. The approach is applied to two constructive forms of the same barrier having increasing complexity. Results are critically discussed by combining complementary remarks on the optimization convergence, the phenomenological reading of the flow around the optimized barrier, and its design and construction. •Aerodynamic shape optimization of barriers against windblown sand is developed.•Barrier trapping performances are maximised, and construction costs minimized.•Computational Wind Engineering simulations are carried out to evaluate the goal function.•Sensitivity of the barrier aerodynamics to design parameters is preliminary assessed.•Performances of both gradient-based and genetic algorithm optimizations are discussed.