In this chapter, we present various optimization results of spatial frames, namely, latticed domes and long-span arches. Following the historical review in Sec. 6.1, we carry out sensitivity analysis and optimization of archtype trusses and a double-layer cylindrical grid in Sec. 6.2, as illustrative examples. In Sec. 6.3, single-point-search heuristic approaches, e.g., greedy method, simulated annealing, and tabu search, are applied to the optimal design of a spatial frame, and their performances are compared. In Sec. 6.4, an approach is presented for incorporating the designer's preference of shape of an arch-type frame that is described using a Bezier curve. Multiobjective shape optimization of a single-layer latticed shell is presented in Sec. 6.5. In Sec. 6.6, a method based on the genetic algorithm is presented for conguration optimization of arch-type trusses incorporating explicit geometrical constraints. In Sec. 6.7, a parametric programming approach is presented for estimating the eect of spatial variation of seismic motions on optimal solutions. Finally, a substructure approach is presented in Sec. 6.8 to optimize a roof truss without carrying out analysis of the whole structure at each optimization step.