<![CDATA[A very poor understanding of mathematics is the driving force behind this research. It is far from the expectations of the achievement of the mathematics learning process, based on the results of the observation of the learning of mathematics in class V. This is because at the time of taking the initial datathat obtained a valueof <75 as much as 100% of all students, two of them obtained a value of 60. Problem Based Learning is a model which, when used as a medium for students to think critically and be skilled at solving problems in order to understand them, contains the characteristics of problem based learning. This study is intended to investigate whether using a model of problem based learning referred to as PBL, students' knowledge of mathematics increases and influences.The experimental research, involving the design of a nonequivalent control group, is based on this research. Where are the 5th grade students as the population in this study, so that two classes are given different treatment as a sample with a random sampling selection technique. Based on the N-Gain value in the experimental class and the control class each get a maximum value of 99 and 90 while the minimum value is 64 and 38 with an average of 79.04 for the experimental class and the control class gets 62.15 where the value indicates the level of data spread to the average value of the data is 9.667 in the acquisition of the deviation value in the experimental class and 11.299 in the control class. It is included in the Effective category with a minimum NGF of at least 64% and a maximum NGF of at least 99%, based on the results of the calculation of the NGF test on the data, which shows an average NGF of 79.04% for the class receiving special treatment. In addition, the average NGain which is defined as a sufficient category of 62.15% while the minimum and maximum gains in this control class were 38% and 90% respectively are set out below.]]>
