Developing a domain-specific language (DSL) or a composition of DSLs to model a system concern, such as deploying and configuring software components to meet real-time scheduling constraints, is time consuming. Ideally, developers should be able to reuse DSLs and DSL compositions across projects to amortize development effort. Reusing DSLs is hard, however, because they're often designed to precisely describe a single domain or concern. A new approach uses techniques from software product lines (SPLs) to improve the reusability of a DSL, DSL composition, or supporting tool by providing traceability of language concepts to DSL design. A case study of four DSLs demonstrates the need for-and benefits of-applying SPL reuse techniques to DSLs. © 2009 IEEE.