Proficiency, which refers to a concept that is more complex than is generally acknowledged, is an overworked term in second language teaching that lacks a satisfactory operational definition. Linguistic knowledge is currently defined most often in terms of what an individual is able to do with the knowledge, as in the movement for competency-based education. The simple idea of proficiency as the ability to perform real-world tasks with a specified degree of skill becomes problematic when the issue of proficiency assessment arises. All proficiency scales assess a mixture of factors from diverse domains. Most of the performance factors have not been empirically validated to determine whether learners really do act in the ways described by the scales or to what degree the skills mastered in one domain are transferable to another. The only performance factors subjected to empirical validation have been syntax and morphology, and it has been found that some aspects of proficiency descriptions conflict with what learners are able to do at other stages. Research into the divisibility of the construct of general proficiency has not been as fruitful as originally anticipated. Despite some valuable research, efforts at producing an operational definition of proficiency seem to be circular and will not be resolved without more empirical study. (MSE)