This study aims to determine (1) the effectiveness of the application of the think pair share (TPS) model on the value of students' nahwu lessons. The method used in this research is quantitative with a Quasi Experiment approach. The subjects in this study were students at the Daeussholah boarding school in Bangkalan. The research sample amounted to 25 students. The instrument used by researchers is a written test. The data analysis technique used paired sample t-test using SPPS version 22. with the hypothesis H0 = there is no improvement from pre-test to post-test results after treatment, and H1 = there is an improvement from pre-test to post-test results after treatment. The results prove that after being given the treatment, namely the think pair share learning model, the average value of the experimental pos-test is greater (85.28) than the average value of the control class pos-test (77.78), through statistica l calculations it is known that the sig. (2-tailed) of the post-test in the experimental class and control class is smaller than 0.05, namely 0.000. The decision made from these results is H0: there is no significant improvement between the think pair share learning model and the qawaid and touch method in nahwu learning. Because, H1 is accepted, so it can be concluded that the think pair share learning model is effective for increasing the value of nahwu lessons for students at Pesantren Darussholah Bangkalan.