The purpose of this study was to investigate the nature of second language proficiency with respect to its divisibility and components. Specifically, the correlated‐traits and the second‐order hypotheses were evaluated using 334 university students learning Englishasa second language, a wide range of language proficiency measures, and confirmatory data‐analytic techniques. The focus was to determine the extent to which each model represented the data. The results of this study provide evidence supporting both models. The significance of these findings as well as directions for future research are disc