One of the problems in the theory of action is that of individuation of action. The difficulty arises when we consider, for example, whether someone's pulling the trigger is identical with his shooting a person. I shall show that to think they are identical, a view proposed by Donald Davidson, must lead to a paradoxical situation. But, on the other hand, to think they are distinct causes another difficulty. Apparently there can be more than one answer to the question "What did he do? ". Thus, both "He pulled the trigger" and "He shot a person" seem to be acceptable answers. But then did he do two distinct actions, for we assume they are not identical? Generally, should we say he did as many actions as there can be answers to the question what he did? This is queer. The purpose of this paper is to point out some of such dilemmatic problems regarding action and to suggest a key to their solution.