This research is motivated by student learning activities that do not use discussion learning models much in Islamic Jurisprudence subjects. This study aims to determine the importance of a learning model Numbered Head Together (NHT) at Cikal Bimbel and analyze the effect of the Learning Model Numbered Head Together (NHT) on the activity and learning outcomes of Class A. Type of research using Mixed Method Research namely mixing Quantitative and Qualitative with design Sequential Explanatory Design where in this study it was preceded by quantitative research and then only complemented by qualitative research. The research sample was 44 students. Consisting of 22 Class A students who received the learning model treatment Numbered Head Together (NHT) and 22 Class B students with the conventional model. Data collection techniques using observation sheets and tests. The data analysis technique used was the Manova Test, which had previously passed the prerequisite tests, namely Normality and Homogeneity. Research results show that students who were given treatment in the form of a Learning Model Numbered Head Together (NHT) were more active compared to students who were not given treatment. The Manova Test of Activeness and Learning Outcomes shows a sig. 0.000 <0.05, then Ho is rejected and Ha is accepted. Based on this, it can be concluded that there is a significant influence on the learning model Numbered Head Together (NHT) on the activity and learning outcomes of Class At Cikal Bimbel. These results indicate that: First, the NHT learning model can in fact be applied to subjects other than exact; Second, teachers get a lot of benefits from the NHT learning model because this model is something new for teachers at Cikal Bimbel, especially in Jurisprudence subjects; Third, the discussion learning model is the basis for active learning so that the teacher must be creative; Fourth, other variables that influence apart from the variables studied are factors from within students and infrastructure.