Three experiments examined the hypothesis that it was easier to process a passage4 if given-information appeared before a new-information in each sentence than in its contrary. In Experiment 1, 20 graduate and undergraduate students were divided into 2 groups and were asked to read each kind of the above passage, sentence by sentence, while using a computer's key. The reading time and the quality and quantity of a free recall, summary, and question answering were compared. In Experiment 2, two groups of 10 subjects were to read each passage forward and backward using 2 keys. The progress of reading and the quality and quantity of cued recalls were compared. In experiment 3, two groups of 10 subjects read each passage within 5 seconds per sentence, and another 2 groups would read within 3 seconds but it was too short a time to read backward. The quality and quantity of cued recalls were compared. All the results of Experiments 1-3 support the hypothsis except the result of question answering in Experiment 1. The results were discussed in terms of the theory of elaboration and a processing model.