Compilation and utilization of second language word lists have the potential to positively or negatively affect curriculum design, materials development, instructional practice, and high-stakes assessments. This study firstly exemplifies this phenomenon by tracing the compilation of the "unauthorized" English language Reference Word List (RWL), the de facto list of words used for Taiwanese high school textbook and materials development compiled by the College Entrance Examination Center. Secondly, this study examines the influence of the RWL and word property variables (Polysemy, Part of Speech, Word Length, and Word Family Size) on non-English majors' (n = 566) vocabulary acquisition (VA). Results show medium to large correlations between RWL Level/Inclusion and VA as well as small to medium correlations between Polysemy, Frequency, Word Length and VA. An eight explanatory variable sequential regression accounted for 50% of the variance in VA, with RWL Level and RWL Inclusion adding the most explanatory power to the model. Next, washback effects of the RWL on Taiwanese high school English learners' VA are discussed and suggestions on compiling an empirically-informed new RWL are provided. Lastly, our critique and suggestions for revision of the RWL are discussed in connection to the creation and revision of word lists for other learning contexts. (C) 2018 Elsevier Ltd. All rights reserved.