Machine tool spindle design is critical for high speed precision machine tools. Dynamic characteristics of spindle, such as spindle speed, torque, dynamic stiffness and thermal characteristics, are influenced by heat generated in the spindle that depends on bearing locations, preload, and bearing geometry. With the increase in bearing preload the spindle stiffness increases. However, heat generation increases with increase in preload, which is detrimental for the bearings. A tradeoff between these two characteristics need to be maintained. A multi-objective optimization formulation was proposed in the present work to optimize the natural frequency and power loss from bearings using thermo-mechanical model of the spindle. Non-Dominated Sorting Genetic Algorithm-II, a multi-objective optimization algorithm, was used towards this for identifying the design parameters of high-speed machine tool spindle. A pareto optimal front was identified which describes optimal tradeoff solutions between the two objective functions.