To help evaluate driver education, 50 individuals were tested and trained by means of a simulation technique to determine if drivers could be taught to make accurate last minute passing judgments, and if knowledge of direction and magnitude of error on individual trials would have any influence on this training. In a projection room, each subject watched as the camera car he was supposedly driving actually passed a lead car. Following this sequence each subject was administered a pretest, involving an oncoming scene film during which the subject was to press a switch when he believed the last possible moment for passing had arrived. Then in followup film sessions one group of subjects received instant feedback on each performance while a second group did not. Analysis of the data showed that there was a significant correlation (.57) between the on-the-road and in the lab performance on this type of task. It was also found that subjects receiving the feedback improved significantly in a posttest measure while the other group failed to improve in accuracy of passing judgment. (Author)