An increasing number of accounting educators are reporting that they use the microcomputer in their classes. Generally, this use has involved templates and has been confined to financial and managerial accounting topics. This study reports the results of an experiment which utilized a student-designed solution to a cost accounting problem. Each of the groups was given a complex process costing problem; one group was required to use an electronic spreadsheet program (LOTUS) to solve the problem, while the other group used the traditional pencil-paper method. Although the LOTUS group improved their performance both absolutely and relative to the pencil-paper group, the difference was not statistically significant. Futhermore, it was found that the attitude of the LOTUS group was not as good as the pencil-paper group with regard to the class or the instructor. © 1988 Taylor & Francis.