We report on a study in which subjects viewed color video stills of natural traffic situations while eye movements were recorded. A display change could occur randomly during three different occlusion modes--blinks, blanks and saccades--or during a fixation. These changes could be either relevant or irrelevant with respect to the traffic safety situation. Furthermore we contrasted insertions and deletions. All occlusion modes appeared equivalent concerning detection rate and detection time, and only differed from the fixation condition. The results also show that the detection of relevant changes was more likely and faster than that of irrelevant ones. However, even relevant insertions, which were almost always detected, were around 180 ms longer to report when they occurred during an occlusion. Furthermore, the detection of relevant changes was fairly stable across a wide range of the visual field, whereas irrelevant changes were less well detected, the further away from the fovea they occurred. We close with an outlook on a follow-up study where only relevant insertions and the blank occlusion were used in a driving simulator environment. Surprisingly, we found an advantage in change detection rate and time with blanks compared to the control condition. Change detection was also good during blinks, but not in saccades.