Object-oriented instructional design (OOID) offers the promise of universal access to online instructional materials, increased productivity among trainers and educators, and solutions for individualizing learning. However, it is unclear whether it can fulfill these promises to the degree many envision. As with every new instructional technology, it is easy to become overoptimistic about learning objects, but problems of education are always more complex than technology alone can solve. In this article, I take a critical look at the proposed benefits of learning objects described in the published literature, particularly scalability and adaptability. I also look at both the difficulties in defining the term learning object and the limitations of metaphors used to describe the concept, and concludes with propositions for learning object usage.