Multiple object tracking (MOT) has proven to be a powerful technique for studying sustained selective attention. However, surprisingly little is known about its underlying neural mechanisms. Previous fMRI investigations have identified several brain areas thought to be involved in MOT, but there were disagreements between the studies, none distinguished between the act of tracking targets and the act of attending targets, and none attempted to determine which of these brain areas interact with each other. Here we address these three issues. First, using more observers and a random effects analysis, we show that some of the previously identified areas may not play a specific role in MOT. Second, we show that the frontal eye fields (FEF), the anterior intraparietal sulcus (AIPS), the superior parietal lobule (SPL), the posterior intraparietal sulcus (PIPS) and the human motion area (MT+) are differentially activated by the act of tracking, as distinguished from the act of attention. Finally, by using an algorithm modified from the computer science literature, we were able to map the interactions between these brain areas.