Abstract: This research attempts to find out the different effects of programmedinstruction and traditional instruction on students' achievement, learningtime, thinking ability and learning motivation. The experiment was carriedout in two first-year classes (an experimental class and a control one) ina middle school, both of which were taught rational number in theiralgebraic course, and by the same teacher. The average level of the twoclasses was approximately equal in the beginning. The experimental classbegan to use a programmed teaching material compiled by ourselves forabout a month, during which both classes were tested a few times and theirmarks compared. The results showed that the experimental class madeobviously better progress in all aspects than the control class. These resultsare probably related to the heuristic characteristics of programmed materials.Programmed instruction still needs further research.

Key words: NULL