The purpose of this study was to determine: (1) Is there any difference between students’ mathematical representation ability who were taught by using the learning model of Project Based Learning with students who were taught by Guided Discovery Learning. (2) Is there any difference in the students' mathematical representation ability based on high, medium, or low learning motivation categories. (3) Is there an interaction between the learning model and the learning motivation toward mathematical representation ability. This research type was quasi experiment. This research was a quasi experimental research with 2x3 factorial design. Instrument used to collect data was test of representation ability and learning motivation questionnaire. The hypothesis testing used two way analysis of variance with unequal cell. Based on the results of hypothesis testing, it can be concluded that (1) There is a difference in the students' mathematical representation ability that was taught by using Project Based Learning model with students who were taught by sing Guided Discovery Learning model with sig value. <0.05 i.e. .000 <0.05 (2) There is a difference in the students’mathematical representation ability based on high, medium, or low learning motivation category with sig value. <0.05 ie .017 <0.05 (3) There is an interaction between the learning model and the learning motivation on the mathematical representation ability with the sig value. <0.05 ie .002 <0.05.