The Involvement Load Hypothesis (ILH) is a framework for designing vocabulary-learning tasks which was proposed by Batia Laufer and Jan Hulstijn in 2001. It assumes that task effectiveness depends on three components induced by a task: a motivational component (need) and two cognitive components (search and evaluation). The hypothesis has been investigated in numerous studies, which have provided varying degrees of support for it. This article reviews the ILH 20 years after it was first proposed and addresses issues arising from the research. The first category of issues is related to the assumptions of the hypothesis, including uncertainty about the relative weight of the components, lack of evidence regarding the effect of the distribution of the components, and the limited range of scores available to assign to tasks. The second category involves variables affecting the predictive ability of the hypothesis, including time on task, level of proficiency, and frequency of exposure. Two studies, which investigated these issues, are discussed. One of them suggested extending evaluation, clarifying the role of search, and redefining need to solve the issues related to the assumptions. The other revealed the effects of the variables in addition to showing the usefulness of the framework for predicting task effectiveness and the relative weight of the components. Issues that remained unresolved are discussed.