Accurate and efficient modeling, optimization, and synthesis of integrated spiral inductors continue to hinder the automated design of mixed-signal circuits in system-on-chip technology. In this thesis, we develop a modeling and automated design methodology for integrated spiral inductors. We have created a wideband inductor model based on closed-form analytical expressions to capture a plethora of resistive, inductive, and capacitive parasitic effects. Leveraging the speed of the inductor model, we have developed a variability-aware automated design methodology that efficiently generates Pareto-optimal inductors based on application requirements. At its core the automated design methodology employs a scalable multi-level single-objective optimization engine that integrates the flexibility of deterministic pattern search optimization with the rapid convergence of local nonlinear convex optimization. The results demonstrate that the inductor modeling, optimization, and synthesis methodology accurately locates and characterizes near-optimal inductor designs with orders of magnitude speed improvement when compared with existing modeling and optimization techniques.