This article describes the development and validation of the new CATSS (Computer Adaptive Test of Size and Strength), which measures vocabulary knowledge in four modalities -- productive recall, receptive recall, productive recognition, and receptive recognition. In the first part of the paper we present the assumptions that underlie the test -- the importance of form-meaning link and the concept of four degrees of knowledge that are implicationally scaled. In the second part of the paper we present the core features of the test in terms of item selection, test format, scoring, and feedback to learners. The third part of the paper describes the process of test validation with 453 students of three English proficiency levels. We adopted the argument-based approach to validation and performed Rasch analyses, Cronbach's alpha, and multiple sets of ANOVAs and Post-hoc tests. The results of these analyses demonstrate the stability, sensitivity, and adequacy of the scoring system, the potential generalizability of the scores, the extrapolation of the scores as indicators of learners' proficiency level and expected vocabulary sizes and the possibility of using the test when making placement considerations or assessing progress.