This study investigated the implementation of Jigsaw Cooperative Learning (JCL) in teaching reading comprehension in one of Junior High Schools in Serang Regency, Banten Province. It seeks to see how the JCL is implemented in the classroom, whether JCL can enhance students’ comprehension and whether the impact is positive or negative in students’ opinion. It employed classroom action research (CAR) with two cycles of actions, involving 31 students. To collect data, observation, tests and questionnaire were done. The implementation shows that, like other cooperative learning models, JCL requires students work in groups twice, in jigsaw and experts groups and require teachers do more preparation. The results of tests indicate that JCL improved students’ comprehension. There was gain of the mean, the highest score and the lowest one. In addition, the data from questionnaire show that in the students’ opinion, JCL impact on their literal comprehension is better than that on their inferential and evaluative comprehension meaning that the students are more aware of the impact on their literal comprehension than that on inferential and evaluative levels. It is suggested that similar study should include more items on every levels of comprehension, involve more students with various ages, be conducted in more cycles and use interview to collect data.