This type of research is Experimental research, which consists of 2 classes, namely the experimental class (the class that learns to use the Discovery learning model) and the control class (the class that learns to use conventional learning models). This study aims to study the differences in learning outcomes between the experimental class and the control class, as well as to learn about the effectiveness of Discovery learning on improving student learning outcomes on Geography subjects at SMAN 8 Makassar. From the results of the study, data obtained showing the results of the learning class increased by 4.34 and the increase in class control increased by 10.12 with an average value of 73.4 and 73.76. While the results of inferential analysis using the t-test show the value of sig. 2 tailed> 0.05, it does not mean a significant difference in learning outcomes of students who learn to use the Discovery learning model with students who learn to use Conventional learning models. The effectiveness of Discovery learning in this study was 43% or included in the category of Effective.