At level 3 automated driving, the driver is out of the loop of the driving task for extended periods during the drive. At system boundaries, the driver must reengage in the driving task and gain situation awareness to successfully take back vehicle control and handle the take-over situation. A lot of development effort is being put on detecting situation awareness with gaze patterns. In the present study, we analyzed eye-tracking data in take-over situations form 31 participants of a driving simulator study on the usage of a level 3 automated driving system. The aim was to identify gaze patterns in the take-over situation that are independent of the driving scenario. We found that gaze patterns in a take-over situation are heavily dependent on the driving scenario. In situations with an acoustic preannouncement, participants glanced at the forward road already before and also after the take-over request. We found that glances to the mirrors were evident only in situations that required a lane change and therefore the visual scanning of side and rear traffic. Only glances to the instrument cluster at the moment of the take-over request were evident for all scenarios. There were no relevant differences in gaze behavior between situations with driving errors and situations without errors. In conclusion, the results suggest that there is no universal glance strategy for building situation awareness in response to a take-over request. Glance pattern seem highly dependent on the requirements of the traffic scenario. Further situational information might be needed to evaluate situation awareness based on gaze behavior.