In this paper, several novel ideas in the definition and optimization of robustness for structures and structural controllers are presented. A robustness bound attributable to Patel and Toda is developed using eigenvalue conditioning analysis. Homotopy and sequential linear programming algorithms are used in lieu of conventional nonlinear programming to implement these ideas for an illustrative example. The numerical results confirm the conservatism of the stability robustness bound but nevertheless support the hypothesis that maximizing the robustness measure does significantly increase the true robustness of a closed-loop system. The results also indicate that maximizing the stability robustness measure produces more robust designs than minimizing eigenvalue sensitivity directly.