The technology acceptance model (TAM) asserts that perceived usefulness (PU) and perceived ease of use (PEOU) are two primary determinants of behavioral intention (BI) and usage. Subsequent studies have mainly focused on the prediction and explanation of behavioral intention to use rather than usage of information systems (IS). This suggests that most researchers embrace, explicitly or implicitly, the notion that if PU and PEOU predict BI, then the two variables must also predict usage. Arguing that BI may not serve as a good surrogate for usage, this study employs a meta-analysis to address an important question: Should researchers concentrate on intention or usage? Our results suggest that BI can not replace usage and PU and PEOU are more likely to predict BI than usage. Based on these findings, we call upon IS researchers to focus their main attention on usage rather than on BI, given that their primary objective is to explain system use and identify its determinants.