Nuclear magnetic resonance (NMR) spectra are among the most common visualizations used by chemists in both academic and industrial settings, where they serve as both representations of submicroscopic entities and tools that support social discourse. The ability to use words to communicate about features within NMR spectra is thus an essential aspect of chemists' representational competence. Collectively, these words constitute a highly specialized language specific to this technique. To adequately prepare chemists, the undergraduate curriculum must therefore cultivate students' ability to use this lexicon. We developed the NMR Lexical Representational Competence (NMR-LRC) instrument, a formative assessment to measure students' ability and perceived ability to use words to communicate the identification, analysis, and interpretation of features within [superscript 1]H NMR spectra. Following development, we administered the NMR-LRC in Spring 2020 to a total of N = 678 second-semester organic chemistry students at a large Midwestern university. We analyzed the responses using the Rasch model to collect statistical evidence of validity and reliability, and we used cluster analysis to evaluate whether the instrument could detect the Dunning-Kruger effect (i.e., students' illusions of competence). Supporting evidence of response process, content, and associative validity was also collected. The results suggest that data obtained from the NMR-LRC can be interpreted as a measure of students' lexical representational competence and associated perceived competence in [superscript 1]H NMR spectroscopy and that the NMR-LRC can detect students' illusions of such competence. Results further suggest that instructors can use the NMR-LRC to formatively assess their students' lexical ability and associated perceived ability following instruction, allowing instructors to provide feedback that promotes students' competence and perceived competence.