This study focused on the use of computer-assisted instruction with high risk students. Subjects were eleventh grade American history students who were presented the same material either by their regular instructor or by a computer sequence developed with an authoring system. Despite a statistically significant advantage for the conventionally-taught students at the time of the pretest, the group receiving computer-assisted instruction (CAI) had a statistically reliable advantage at the time the lessons concluded. It is suggested that this difference can be attributed to three factors: time factors involved in learning the lessons; the fact that students in the conventionally taught group could not make up work missed, but CAI group students could; and a Hawthorne effect whereby the improvement was due to a change in classroom regimen. Thus, it is suggested that CAI offers a potential for remedial instruction for the high risk student, but such CAI should be instructor-designed in order to maximize individualization of instruction. (EW)