Abstract: A genetic algorithm based optimization scheme was proposed to reduce the cost of cable dome structure in practical engineering. The prestress coefficient and member sections with specified profiles were defined as discrete design variables, and structural tensioning level was considered in the object function. Taking Geiger and Levy cable domes as two examples, the minimum cost design was obtained which satisfied constraints of structural stress and displacement in a variety of load conditions, and sensitivity analysis was adopted to study the influence of prestress on optimization results. The optimization process was also carried out for an improved structural plan, ridge-bar cable dome, in which the ridge cables were replaced with steel bars capable of bearing compression, and the structural cost and mechanical properties of improved plan were compared with the original cable dome structure. Results show that the optimization method can gain the minimum cost design of cable dome effectively and accurately,and the internal force of all members are closed to their limit state. To guarantee the structural stability and bearing capacity of different members, it is necessary to consider the partial factor of prestress and non-uniform distribution of live load in the design process of cable domes. Adopting steel bars instead of ridge cables can slightly lower the total cost of cable dome and significantly reduce the deflection and reaction force.