Over the past two decades, the Involvement Load Hypothesis (ILH) has become a popular buzzword in the field of Second Language Acquisition (SLA). Although applications of the ILH can improve students' learning of productive vocabulary, this effect appears to be transitory. Students' learning of productive vocabulary often fades over time, as shown in delayed productive vocabulary tests. To address this problem, the present study investigated the effect of supplementing the ILH with vocabulary-use knowledge on students' productive vocabulary performance. We designed two WeChat applets, namely Applet 0.0 and Applet 1.0. The design of Applet 0.0 was based solely on the ILH, whereas Applet 1.0 was guided by the ILH supplemented with vocabulary-use knowledge, with the latter consisting of grammatical functions (the grammatical patterns that a word can fit into), collocations (other words that appear together with a word), and constraints on use (the situations in which a word is used). Fifty-one English foreign language undergraduates were assigned to a control group (CG) or an experimental group (EG). Both groups completed a weekly paragraph-writing task as they learned vocabulary with the applets (the CG used Applet 0.0 and the EG used Applet 1.0). The results suggested an advantage for the ILH supplemented with vocabulary-use knowledge over the ILH alone for productive vocabulary performance. Moreover, mobile-assisted language learners' language proficiency levels did not influence the effectiveness of the ILH (with or without vocabulary-use knowledge) on productive vocabulary performance. The EG students' perceptions of the WeChat-mediated Mobile-Assisted Language Learning approach are also discussed.