Virtual reality (VR) holds great promise for the study of perception-action. The case of studying the outfielder problem is presented as an example of how VR has contributed to our understanding of perception-action, and of the potential and pitfalls of using VR in such a task. The outfielder problem refers to the situation in a baseball game (and analogous situations) in which an outfielder has to run to get to the right location at the right time to make a catch. Several experimental studies are discussed in which participants had to intercept real or virtual balls. The biggest added value of using VR is the fact that the virtual world is completely in the hands of the experimenter, which allows studying situations that do not exist outside of VR, thus enabling strong hypothesis testing. A number of factors related to the success of the VR experiments are identified, such as the lack of haptic feedback in VR setups used in this paradigm until now, the specifics of the optics presented to the participants, and the available space for locomotion. We argue that it is important to make a close comparison of task behavior in VR with that outside of VR, but conclude having great expectations of the role of VR in perception-action research.