Although the genetic algorithm(GA) for structural optimization is very robust,it is very computationally intensive and hence slower than optimality criteria and mathematical programming methods.To speed up the design process, the authors present an adaptive reanalysis method for GA and its applications in the optimal design of trusses.This reanalysis technique is primarily derived from the Kirsch's combined approximations method.An iteration scheme is adopted to adaptively determine the number of basis vectors at every generation.In order to illustrate this method, three classical examples of optimal truss design are used to validate the proposed reanalysis-based design procedure. The presented numerical results demonstrate that the adaptive reanalysis technique affects very slightly the accuracy of the optimal solutions and does accelerate the design process, especially for large-scale structures. Although the genetic algorithm (GA) for structural optimization is very robust, it is very computationally intensive and therefore slower than optimality criteria and mathematical programming methods. Speed ​​up the design process, the authors present an adaptive reanalysis method for GA and its applications in the optimal design of trusses. This reanalysis technique is derived from the Kirsch's combined approximations method. An iteration scheme is adopted to adaptively determine the number of basis vectors at every generation. In order to illustrate this method, three classical examples of optimal truss design are used to validate the proposed reanalysis-based design procedure. The presented numerical results demonstrate that the adaptive reanalysis technique affects very slightly the accuracy of the optimal solutions and does accelerate the design process, especially for large-scale structures.