The present study was undertaken in the hope that our Reading Accelerator might promote reading rate and comprehension, especially the former. Our Reading Accelerator used in this study was made on the model of the Reading Accelerator by Science Research Associates of Chicago and was adapted to suit reading condition of this country. A moving curtain, installed on the Reading Accelerator, hides the plane of this instrument by covering parts of it at the definite rate, which the Ss are able to adjust in ten speed steps. One hundred and one 5th grade schoolchildren of two elementary schools in Tokyo were devided into five groups (A3, B3, Al, Bl and C), which were equivalent in terms of intelligence and reading abilities. Before and after practice sessions, which were about three months long, each subject was tested for his reading rate and comprehension. Group A3 (18 children) and Group Al (13 children) practiced with the instrument, the former practiced three times a week, the latter once a week. Group B3 (17 children) and Group Bl (15 children) practiced without the instrument, the former practiced reading three times a week, and the latter once a week. Group C (38 children) had no experimental reading. All Ss except Group C were asked to read silently a mimeographed story twice. They were askied to try the Outline-Comprehension-Test after the first practice of reading and to try the Detailed-Compehension-Test which contain 10 questions after the second practice. During each practice the experimenter measured S's reading rate and comprehension. Experiments were done individually rather than in groups. Each subject had eight practice sessions. In each session he read one of eight stories, which were nearly equal in difficulty. The difficulty of stories was examined by means of the "Cloze" procedure. To compare eye-movement in reading with and without the accelerator, the eye-movements of one subject, a 5th grade child, were recorded by an electro-oculographic method (or Galvanometric method) in both conditions. Results : 1. During the practice period the Ss' reading rate made progress on the whole though the differences between the four groups were slight. Increases in reading rate of Group Al and Bl were the largest. 2. When test scores before and after practice sessions were compared, the following results werer found. The Practice Group (Group A and B) lead in the gains made in reading rate when compared with the Group C. 3. In reading rates group Al and Bl took a pronounced lead over Group A3 and B3 who had practiced three times a week, showing that Al and Bl training was more effective thatn A3 and B3 training. 4. In comprehension score no significant difference was found among groups. 5. In comparing Group Al with Group Bl, progress in the gain of Group Al who had practiced with the Reading Accelerator was more remarkable than that of Group Bl, without the instrument. 6. Therer was no significant difference between the gain of Group A3 and that of Group B3. 7. The eye-movements in reading with the Reading Accelerator were more smooth than free-reading without the instrument. There were fewer stops with the Reading Accelerator. These results lead us to the conclusion that one can accelerate one's reading rate in a few months if one practices reading with our Reading Accelerator, and that the distributed practice is more effective than the mass practice.