In investigating satisfaction drivers which involving latent variables, the application of Structural Equation Model or LISREL. many use CB-SEM as the parameter estimation approach. However, many researchers do not pay enough attention to the adequacy of the data in meeting the estimated requirements with the Maximum Likelihood method to obtain a suitable solution in LISREL modeling. PLS appears as an alternative to SEM analysis that does not require a large sample size to produce the proper solution. The purpose of this study is to examine the factors driving satisfaction in sharia banking using PLS and SEM. The satisfaction model used involve latent variables which is Islamic Product Attributes, Religious Commitments, Service Quality, Trust, Satisfaction and Loyalty. Data was collected through the results of answers to questionnaires that used a Likert scale to 209 Bjb sharia bank customers. The results of the comparison of the use of these two approaches show that the estimation result of CB-SEM loading factor is higher than PLS-SEM, while the estimation for the path coefficient with PLS-SEM is higher than CB-SEM. By using both methods, the most important and significant factor in increasing the satisfaction of sharia bank customers is Trust.