<![CDATA[ABSTRACT The use of fewer media to explain material resulted in students is less active when the learning process so that students have difficulty understanding material. This is guessed to be one of the factors of mathematics student’s value is low. This research is intended to know (1) whether or not differences in mathematics learning outcomes by using GeoGebra assisted problem-based learning model with a direct learning model, and (2) to know the effectiveness of two models on the learning outcomes of X class Of SMA Muhammadiyah 5 Yogyakarta In Academic Year Of 2016/2017. The population of this research is students of X classes of Muhammadiyah Senior High School 5 Yogyakarta (SMA Muhammadiyah 5 Yogyakarta) in the academic year 2016/2017. Samples were taken by random sampling technique to class, to get XF as experiment class and XB as control class. This experiment design is a post-test only control design. The data get from the post-test result in the description. Analysis of test instrument use validity test and reliability test. The analysis technique consists of (1) test of requirement: test of normality and test of homogeneity, and (2) test of hypothesis. Test of hypothesis use statistics test of t-student. Hypothesis testing is done two times; the first hypothesis and the second hypothesis. The result on significance is 5%, and df = 62 showed that: (1) there were differences in mathematics learning outcomes using GeoGebra assisted problem-based learning model with a direct learning model. It is showed by tcount = 5,031541 and ttable = 1,999. So that tcount = 5,031541 > ttable = 1,999, (2) problem-based learning model more effective than direct learning model to learning outcomes of the student. It is showed by tcount = 5,031541 and ttable = 1,670. So that tcount = 5,031541 > ttable = 1,670.]]>
