The purpose of this study is to determine the effect of the TGT (Teams Games Tournament) learning model on statistics material for class VII junior high school on mathematical problem solving abilities. This research was a quantitative research using Quasi Experimental Design. The sampling technique in this study used purposive sampling. Class VII D, consisted of 24 students as the experimental class and class VII C consisted 24 students as the control class. The instruments used in this research were Pretest-Posttest with 3 questions. The data analysis techniques used in processing the data were analysis prerequisite tests (normality test and homogeneity test), hypothesis testing, and the N-Gain test. Based on the results of calculations that have been carried out, the posttest average value for the experimental class was 92, while the posttest average value for the control class was 85. In addition, based on the data obtained from the results of the hypothesis testing, it results that the value is 0.000 then Ho is rejected and Ha is accepted. This is in line with the average N-Gain value in the experimental class resulting in 0.82 high criteria, while the average N-Gain value in the control class produces 0.66 medium criteria. So it can be concluded that there is an influence of the TGT (Teams Games Tournament) learning model on students' mathematical problem solving abilities.