Global robust stabilization is considered for a class of uncertain nonlinear systems and a new systematic tool for robust backstepping is developed. By choosing state-dependent scaling functions in each step of the recursive design, one can construct robust control laws for an extended class of strict feedback systems. The scaling approach to robust backstepping treats both static and dynamic uncertainty in an unified way. The method is amenable to automatic computation based on numerical optimization, and for instance, one can thus introduce performance objectives such as instantaneous control effort minimization into the design.