Previous studies have proved that there is a highly localized object-based inhibition on nontargets in Multiple Object Tracking (MOT) task. And the mechanism of selective nontarget inhibition was closely related to the difficulty level of tracking task. When there’s a high demand for tracking task, more inhibition was obtained to the nontargets that were hard to distinguish from targets within the capacity of Visual Index. Using Multiple Object Tracking with Probe Dot Detection task, the present study further examined the effects of Multiple Object Tracking difficulty on the mechanism of nontargets inhibition. Two experiments were conducted and both were 3×2 within-subject design. The two independent variables were moving nontargets’ number (experiment 1: having 4, 5 or 6 moving nontargets while there were always 4 targets and 4 static nontargets) or static nontargets’ number (experiment 2: having 4, 5 or 6 static nontargets while there were always 4 targets and 4 moving nontargets) and probe dot locations (probe dot presented in static or moving nontargets). Participants in two experiments performed no-tracking task first and then the tracking task. In no-tracking task, participants only need to monitor if a probe dot were presented in the objects. In tracking task, participants must track the targets while monitoring the probe dot. Participants’ tracking accuracy and probe detection performance on static and moving nontargets were recorded. And the degree of inhibiton on static and moving nontargets was obtained through percentage of probe detection in no-tracking task subtracting that of tracking task. The results showed that in two experiments, when the number of moving nontargets (experiment 1) or static nontargets (experiment 2) was 6, the tracking task was significantly more difficult, and the degree of inhibition on static nontargets was significantly larger than that of 4 and 5 moving or static nontargets. However, the inhibition on moving nontargets kept constant among the three conditions. And in experiment 2, when the number of static nontargets was 6, the degree of inhibition on static nontargets was significantly larger than that of on moving nontargets. In general, the findings of the present study suggested that in Multiple Object Tracking task the inhibition on static nontargets was sensitive to the difficulty level of tracking task. The degree of inhibition on static nontargets increased with the increase of difficulty level of the tracking task. While the inhibition on moving nontargets kept constant among different difficulty levels. And when the number of visual objects exceeded the capacity of Visual Index, in order to complete the tracking task, participants would change the mechanism of selective nontarget inhibition, the inhibition on static nontargets would increased.