In evolutionary computation, crowding is a popular technique to handle multi-modal optimization problems, which include many possible local or global optimal solutions. In our previous publication, we proposed a new evolutionary algorithm, called DEAL (Direction-guided Evolutionary Algorithm). It works effectively on non-linear optimization problems. In this paper, we extend further DEAL towards the area of multi-modality by applying a crowding mechanism, called as CrowdingDEAL. We validated CrowdingDEAL algorithm with a wide range of benchmark problems. The obtained results indicated a strong performance of CrowdingDEAL in dealing with multi-modality and in comparison with other algorithms.