Object individuation refers to the task of diciding the number of distinct objects present in a given scenario. A central issue within the field of object individuation concerns the kind of information infants rely on when they individuate objects successfully. By means of the violation-of-expectation strategy, the present study reports a comparison of 8.0-, 6.5-, and 5.0-month-old infants' use of featural and spatiotemporal information in an event monitoring design. Using a mirror setup the memory demands was minimized, because all apparent changes in the unexpected test events took place in full view of the infants. The results indicate that while 8.0-, and 6.5-month-old infants individuated objects successfully regardless of whether they were provided with featural or spatiotemporal information, only the 5-month-old infants provided with spatiotemporal information succeeded. The results are discussed in relation to the relevant litterature.