This study reviewed use of covariance-based structural equation modeling (SEM) in 209 articles published in nine tourism journals between 2000 and 2011. Time-series analysis suggested that the number of SEM publications is explained by linear and quadratic time effects. Results indicated that although SEM practices have improved in some areas, tourism researchers do not always engage in the recommended best practices. Problematic areas were related to testing of alternative a priori models, reporting of multivariate normality, estimating effect size, and assessing reliability and validity. Some of the recommended fit indices were also underutilized. The review also found few differences in SEM practices between top-tier journals and other publications in the tourism field. On the basis of these results, suggestions to improve use of SEM in tourism studies are discussed in the light of recent developments in this technique.