The popular matching problem introduced by Abraham, Irving, Kavitha, and Mehlhorn is one of assignment problems in strategic situations. It is known that a given instance of this problem may admit no popular matching. For coping with such instances, Wu, Lin, Wang, and Chao introduced the popular condensation problem whose goal is to transform a given instance so that it has a popular matching by deleting a minimum number of agents. In this paper, we consider a matroid generalization of the popular condensation problem, and give a polynomial-time algorithm for this problem.