This study aims to analyze the ability to solve mathematical problems based on the characteristics of students' thinking style: Concrete Sequential (CS), Abstract Sequential (AS), Concrete Random (CR), and Abstract Random (AR). This research was carried out at one of the junior high schools in Tangerang, Indonesia. The subjects were 20 Year 7 students, with different thinking style. This research employed a qualitative approach. The instruments used are problem-solving tests and non-problem-solving tests on the characteristics of students' thinking style and a test of mathematical problem-solving ability accompanied by interview results. The data were analyzed descriptively to describe the test results of students' mathematical problem-solving abilities. The results showed that CS-type students could understand the problem in accordance with the indicators. Type AS students can understand but do not write the steps in detail. AR students are as well-understood but less thorough about their calculations. Meanwhile, CR students can understand only a few parts of problem-solving.