Uniting two areas of current interest in educational research, a project tested the effectiveness of computer software in observing the process of comprehension monitoring. Passages with embedded comprehension obstacles were presented to two groups of fifth grade students. Identical in content to those of the controls, the experimental subjects' passages flashed comprehension strategies on the computer screen when the subjects reached an obstacle. A follow-up comprehension test revealed no difference in scores between experimental or control groups, indicating that the training did not improve comprehension skills. Reasons for the lack of significant findings might include the high degree of motivation of control, as well as experimental, subjects; the insufficient difficulty of the passages; and the failure of students to grasp the purpose of the experiment. The process data collected during students' reading, however, showed that the experimental subjects used a greater variety of comprehension strategies than did the controls. Further research comparing microcomputer-assisted instruction with traditional classroom instruction is now in progress. (MM)