Emerging technology has changed the focus of corporate learning systems from task-based, procedural training to knowledge-intensive problem-solving with deep conceptual learning. In addition, the deployment of open systems and distributing processing are adding new stresses to learning systems that can barely keep pace with the current rate of change. Learning environments to address these challenges a reviewed within a framework of the conventional learning curve, in which different learning elements are required to support different levels of expertise. An adaptive development model for creating and sustaining a learning environment is proposed that consists of the iterative application of three phases: (1) analysis and reflection, (2) architecture inception and revision, and (3) alignment. The model relies on the notion that analysis deals as much with synthesis and learning as it does with decomposition. We conclude that the concept of a “learning environment” provides a viable construct for making sense of the array of systems designed to support knowledge management, document management, e-learning, and performance support. A learning environment with a well-defined architecture can guide the convergence of multiple systems into a seamless environment providing access to content, multimedia learning modules, collaborative workspaces, and other forms of learning support. Finally, we see future learning environments consisting of networks of databases housing content objects, elegant access to the content, ubiquitous virtual spaces, and authoring tools that enable content vendors, guilds, and universities to rapidly develop and deliver a wide range of learning artifacts.