The purpose of this paper is to investigate students' engagement in computational thinking (CT) and programming with MATLAB when solving a mathematical task in a programming course at the undergraduate level. The data collection method is participant observation of three groups of three students presented with a mathematical task to solve. The study uses a deductive-inductive analysis strategy based on the interplay between the theoretical framework and the empirical data. The results reveal that engaging students in CT for mathematical problem-solving is challenging for many reasons: Lack of mathematical thinking (MT), insufficient experience with CT, and more importantly lack of deeper connection between CT and MT. Conclusions are drawn from the results to promote mathematical problem-solving by means of CT and MT at the undergraduate level. [For the full proceedings, see ED621108.]