The main purpose of the study is to investigate the effects of prospective teachers' cognitive styles on learning approaches. It is aimed to define whether exist significance differences between defining prospective teachers' cognitive styles and learning approaches and demographic variables within the scope of the mean purpose. The study, designed according to correlational survey model, was conducted at Mugla Sitki Kocman University, Faculty of Education in the 2014-2015 academic year spring semester. As data collection instruments, "The Group Embedded Figures Test" was administered to define prospective teachers' cognitive style in the study. On the other hand "The Revised Two Factor Study Process Questionnaire" was used to reveal prospective teachers' learning approaches. According to the findings, prospective teachers generally have field dependent cognitive style. It is determined that between with prospective teachers' gender and academic success and cognitive style scores there is not any significant difference revealed. However, there is significant difference between branches and cognitive style scores. It has been viewed that prospective teachers prefer deep learning approach generally. There is not significant difference between gender and learning approaches yet there is significant difference between learning approaches-branches and academic success. It is also concluded that as prospective teachers' cognitive styles approaches to field independent, deep learning approach preference of prospective teachers has diminished. [This paper was presented at the International Eurasian Educational Research Congress (3rd, Mugla, Turkey, May 31-June 3, 2016).]