The number of student achievements in competition is one indicator to see the quality of learning in a school. This study aims to see the extent to which the application of cooperative learning in improving students' ability to face physics olympiad and other competitions. The sample consisted of 12 students who were prepared to compete and were taken from 3 different grade levels namely 10th grade, 11th grade, and 12th grade in one of the State Madrasah Aliyah (MAN) in Bandar Lampung City. The 12 students were divided into 2: 6 students in the experimental class and 6 in the control class. The study used a type of Quasi Experiment with the design of One Group Pretest-Posttest Design. Learning is done by applying three types of cooperative learning, namely Student Team Achievement Division (STAD), Jigsaw II type, and Teams Games Tournament (TGT) type with a total of 24 meetings. The results of the learning process are obtained from the results of the pre-test given to all students. In the pre-test, it was divided into 2 types, namely: Pre-test grub and Individual Pre-Test. Pre-Test grub is a test that answers questions quickly, precisely, and carefully. In the Pre-test grub, it is done by dividing 12 students into 4 grub (3 students/grub). Pre-Test grub is a simulation of LCT (Lomba Cepat Tepat). Pre-Test grub is an application of cooperative learning method type (TGT) which aims to train speed in answering questions correctly. Individual Pre-Test is Pre-Test which is done by giving students a question sheet to answer. Both types of Pre -Test are given 120 minutes each. After doing the Pre-Test, students test results were evaluated to find out the deficiencies in understanding. Furthermore, the re-learning process is given again in the form of repetition and reinforcement of material that is poorly understood before the Post-Test (grub and individual) is carried out. The results showed that there were differences in understanding the theory and test scores based on differences based on class level, the number of participants in learning activities, and the number of students practicing answering questions. The results of this study were then correlated with their achievements from the results of physics competitions at various events. The results showed that cooperative learning methods with 3 types, namely: type (STAD), type Jigsaw II, and type (TGT) can improve achievement in physics competitions. The conclusions of these results, obtained from the results of the scores on the grub or LCT competitions conducted at different times and places, and also from the results of the comparison of physics olympiad scores (individual competition). The results of this study show that the cooperative learning model has a positive impact on student achievement.