Computational infrastructures provide the framework in which computing can support a particular application. A good infrastructure can be critical to a project as it determines how a program (or set of programs) looks, feels, and is amenable to extension and change. Problem solving environments (PSEs) are a form of computational infrastructure that seeks to provide an integrated set of tools for a particular application area. In this talk, I discuss our recent research and development of component-based PSEs for biomedical computing. Specifically, I will discuss the BioPSE and SCIRun2 problem solving environments. BioPSE is a component-based visual problem solving environment (PSE) that is designed specifically to address large-scale computational problems in biomedicine. SCIRun2 is an evolving PSE based upon the idea of meta-components that “bridge” easily to other component models, allowing one to integrate other existing programs into the shared SCIRun2 framework. BioPSE and SCIRun2 support the entire life cycle of scientific applications by allowing scientific programmers to quickly and easily develop new techniques, debug new implementations, and apply known algorithms to solve novel problems. BioPSE and SCIRun2 also contain many powerful visualization algorithms for scalar, vector, and tensor field visualization, as well as image processing tools (such as ITK). To illustrate the feasibility and utility of computational infrastructures, I will provide examples of several driving biomedical research applications in cardiology, neuroscience, and medical imaging.