Design has traditionally involved selecting a suitable material for a given application. A materials design revolution is underway in which the classical materials selection approach is replaced by design of material microstructure or mesostructure to achieve certain performance requirements such as density, strength, ductility, conductivity, and so on. Often these multiple performance requirements are in conflict in terms of their demands on microstructure. Computational plasticity models play a key role in evaluating structure-property relations necessary to support simulation-based design of heterogeneous, multifunctional metals and alloys. We consider issues related to systems design of several classes of heterogeneous material systems that is robust against various sources of uncertainty. Randomness of microstructure is one such source, as is model idealization error and uncertainty of model parameters. An example is given for design of a four-phase reactive powder metal-metal oxide mixture for initiation of exothermic reactions under shock wave loading. Material attributes (e.g. volume fraction of phases) are designed to be robust against uncertainty due to random variation of microstructure. We close with some challenges to modeling of plasticity in support of design of deformation and damage-resistant microstructures.