This paper describes the development and evaluation of a computer-aided learning (CAL) program. The program was tested in a trial that involved 36 people with diabetes; 20 received CAL lessons in diabetes management and 16 attended conventional diabetes classes conducted by diabetes educators. When measurements taken before and three months after the education were compared, both groups showed significant improvement in their knowledge; the blood glucose levels of the CAL group were significantly lower but those of the conventional education group were higher. This means that the CAL program was as effective as conventional education in imparting knowledge but it was more likely to motivate people to control their glucose levels. The CAL program allows diabetes educators to spend less time on education in basic knowledge and to concentrate more on motivational and social factors that are important determinants of patient compliance. It can also benefit people with diabetes whose access to health professionals and/or conventional diabetes education is restricted.