In many situations, it is necessary to predict when a moving object will reach a given target even though the object may be partially or entirely occluded. Typically, one would track the moving object with eye movements, but it remains unclear whether ocular pursuit facilitates accurate estimation of time-to-contact (TTC). The present study examined this issue using a prediction-motion (PM) task in which independent groups estimated TTC in a condition that required fixation on the arrival location as an object approached, or a condition in which participants were instructed to pursue the moving object. The design included 15 TTC ranging from 0.4 to 1.5 s and three object velocities (2.5, 5, 10 deg/s). Both constant error and variable error in TTC estimation increased as a function of actual TTC. However, for the fixation group only, there was a significant effect of object velocity with a relative overestimation of TTC for the slower velocity and underestimation for the faster velocity. Further analysis indicated that the velocity effect exhibited by the fixation group was consistent with participants exhibiting a relatively constant misperception for each level of object velocity. Overall, these findings show that there is an advantage in the PM task to track the moving object with the eyes. We explain the different pattern of TTC estimation error exhibited when fixating and during pursuit with reference to differences in the available retinal and/or extra-retinal input.