Two projects that have explored the feasibility of computer-generated verbal score reports for the Iowa Tests of Basic Skills are reported. The first, the Madison Project, designed, produced, and evaluated computer-generated testing reports that were useful to teachers and informative to parents. The reports, which were narrative in format, were structured as follows: the first five paragraphs described the pupil's performance on the Iowa Tests of Basic Skills as a whole and on each of the sub-tests; then, lists of specific areas were printed for which the pupil was diagnosed as either proficient or deficient. The teacher report was very similar to the parent report with the addition of imbedded percentile ranks. A class-summary report for the teacher was also generated. This report discussed the class as a whole and concluded with specific suggestions to the teacher for improving class performance in which their performance was weakest. Narrative testing reports were generated for a random half of 52 fourth-grade classes in 17 schools. The greatest shortcoming of these reports was the absence of a pupil report. In the second project, The Iowa Project, which developed a model for a narrative reporting system for the fifth-grade level Iowa Tests of Basic Skills, the most important changes from the Madison Project were: (1) the addition of a pupil report; (2) the inclusion of growth statements; and (3) the use of three alternative levels of the parent and pupil reports--levels that are based on overall pupil performance on the test battery. Examples of the parent and teacher reports from the two projects are provided. (DB)