Abstract: The literature review was used to assess the application of PLS-SEM (partial least squares structural equation modeling) multivariate data analysis on sports spectators.The findings include:Adopting mediating variables and moderating variables is helpful for understanding the relationship between spectator consumption variables. It is important to distinguish the differences between informative variables and reflective variables. Compared with CB-SEM (covariance-based structural equationmodeling), PLS-SEM has the advantages of small sample sizes, non-formal distributions, multiple-variables in complex models and informative measurement models. It is advisable to choose suitable indicators in accordance with latent variables. PLS-SEM is better applied when using small sample sizes, formative indicators included or conducting exploratory research on spectator fields.