This paper presents an optimization model for improving stability levels of thin– walled composite beams under axial compressive loading. Optimum designs are obtained by maximizing the critical buckling load while maintaining the total structural mass at a prescribed value equals to that of a baseline design. The dual problem of minimizing the totaL structural mass under preserved buckling load is also addressed. The developed optimization models deal with slender beam–columns that are axially graded in both material and wall thickness. The main structure is constructed from uniform segments that are fabricated from a composite with different volume fractions of the constituent materials, making the physical and mechanical properties change piecewisly in the axial direction. Design variables include the volume fraction of the constituent materials, the wall thickness as well as the length of each segment composing the beam. The buckling load analysis is performed via finite element method, using a beam element with two degrees of freedom at each node. The resulting optimization problem has been formulated as a nonlinear mathematical programming problem solved by invoking the Matlab optimization toolbox routines, which implement the method of sequential quadratic programming interacting with the associated eigenvalue problem routine. The proposed mathematical models have shown that the use of material grading concept can be promising in raising stability boundaries without mass penalty and producing economical designs having enhanced stability as compared with their corresponding baseline designs. Finally, the given approach can be beneficial to guide structural engineers for choosing the significant design variables in proper and efficient way without violating economic feasibility requirements.