This qualitative case study was conducted to analyze students’ cognitive modeling routes based on their mathematical thinking styles. The participants were 23 tenth-grade students from a school in Iran, six of whom were selected after their mathematical thinking styles had been identified and assigned to three pairs, each representing a different thinking style, for an in-depth analysis of their cognitive modeling routes. Data were collected using three mathematical tasks designed to identify students’ mathematical thinking styles and one mathematical modeling problem. The data were analyzed using the performance characteristics associated with the steps of Borromeo Ferri’s cognitive modeling cycle. The findings revealed that students’ cognitive modeling routes differed according to their mathematical thinking styles. These differences were evident in the sequence in which they moved through the steps of the modeling cycle and in the types of representations they used during problem solving. Moreover, focusing on key information, eliminating irrelevant information, selecting appropriate representations, and engaging in ongoing validation contributed to greater accuracy and efficiency in the modeling process. These findings may inform the design of mathematical modeling activities in mathematics curricula.